These are prospects who’ve been on your website, bought from you before, know someone you know, or are fed up with what they have today. These plays don’t need a big budget, but you have to run them all the time, not just once a quarter when the pipe is light.

What all of these plays have in common is a signal: a job change, a mention on a call, a mutual connection, an event. In the age of AI slop, the teams that win aren’t sending more messages; rather they are catching more signals and acting on them while they’re fresh.

1. Go after the warm leads already on your website

The people visiting your website are already researching your solution, but most won’t fill out a form. Tools like Vector de-anonymize your web traffic and match visitors to your ICP, so you can see which prospects are already on your site. For developer and open source products, tools like Reo.dev and Scarf can show you who is looking at your docs and possibly using the product.

Unless the lead has previously converted on your site, you will likely only have the company data, which is still valuable. For example, at MongoDB we identified the companies downloading the open source software so that a rep could go to a CTO and say, Did you know your team has downloaded MongoDB fifty times this month? We should talk.

2. Use industry events as an excuse to reach out

You don’t need to spend on an expensive booth or sponsorship to get prospects at industry events. You just need to use the event as a reason to reach out and talk to someone.

Ahead of the event, mine the speaker and sponsor list for prospects. As you get closer to the event, conferences often provide an app for networking that you can use to figure out who will be in attendance.

But you don’t need to know that the person is attending. Reaching out to say, “Hey, will you be at AWS re:Invent this year?” is a natural, low-pressure reason to start a conversation, and prospects are more likely to respond, whether they are attending or not.

3. Follow your champions when they move

Your best customers have champions: the people who brought you in, fought for the budget, and made the project a success. When one of them changes jobs, they’re walking into a new company looking for quick wins. Tools like UserGems, Spear, and others will watch for this for you. They flag the moment a champion or past buyer starts a new role, so your message is one of the first they get. It’s another great reason to reach out and start a conversation.

4. Put your network (and especially your investors) to work

People are much more likely to respond when someone they know is vouching for you. So before you cold call, look for a warm intro through your board, investors, advisors, and employees. Don’t stop at your own network: your colleagues know people you don’t.

And make your investors work! Their interests are aligned with yours and they usually have big networks. Make it dead simple for them by sending a short, personalized email they can forward.

Once you’re at scale, tools like Spear and Vieu can do the mapping for you, scanning your whole team’s and your investors’ LinkedIn networks against your target accounts and finding the warmest path into each one.

5. Revisit closed-lost deals and the inbound graveyard

You met with a prospect, they were interested, and the timing was wrong. For the love of God, please keep in touch with them! Budgets and priorities change, so follow up on a regular cadence.

Most companies also have inbound leads that never converted: webinar and event attendees, content downloads, demo requests. Re-engaging these on a regular cadence is waaaay easier than starting with a cold lead.

This is where most teams drop the ball, because nobody owns the follow-up. Treat every moment as a reason to reach out: a birthday, a promotion, a new job, a funding round, a post they wrote. A short, human note on these occasions is an easy way to re-engage leads.

6. Get hyper-targeted

Figuring out a company’s tech stack or key initiatives is easier that you think. Just look at the company’s websites, job listings, and people profiles. With tools like Sumble and Rox, you can now do that at scale to run highly targeted campaigns. For example, you can use natural language queries like “find me companies with over 1000 employees who are using Apache Kafka in the east coast of the US” to build lists of accounts to target and create a message specifically for that audience and their pain points.

7. Run competitive plays

Companies that have already bought a solution in your category both understand the problem and have the budget for it. Look for public references from your competitors: case studies, logo walls, conference talks. The same tools mentioned in #6 (Sumble and Rox) apply here.

Then listen for frustration. Review sites like G2 are a goldmine. If a customer writes a two-star review about the exact thing your product does well, that’s a conversation worth starting.

The most honest complaints are on Reddit. Buyers vent there in a way they never would on G2: “Has anyone actually gotten X to work?” or “Looking for an alternative to Y, it’s killing us.” Set up alerts on your competitors’ names, your category keywords, and phrases like “alternative to” and “switching from.” Watch the subreddits your buyers live in, plus Hacker News, X, LinkedIn comments, and community Slacks.

When you find a thread, don’t drop a sales pitch in it. Redditors can smell that from a mile away. Answer the question, be useful, and then reach out to the person directly with something specific to what they wrote.

8. Find a personal connection (and maybe send a gift)

As I’ve said before, people buy from people. If you’re trying to get in front of a prospect, find out what they care about outside of work: their alma mater, a hobby, a charity, a sports team.

A real example: some of you know I have an alter ego who loves endurance sports. For several years I’ve done a fundraiser for the Mohonk Preserve called Rock the Ridge. While I was at MongoDB, Pocus was prospecting us. The rep didn’t just donate to my campaign page. She also sent me a package of running gear to prep for the race: a water bottle, salt pills, electrolytes, and elastic shoelaces. I still remember it, and still use those shoelaces.

9. Put your executives out front, particularly for small events

A message from a CEO or VP lands differently than a message from an SDR. Executives get replies because buyers want to talk to peers, not to a sequence.

If you host dinners, roundtables, or webinars, the hardest part is getting the right people in the room. You need someone to draw people in, such as a customer, analyst, or expert in your field. When you pair that with getting your executives to invite people personally, it’s much easier to fill these events than sending a marketing email or posting on LinkedIn.

The problem is getting executives to take the time to do that personalized outreach. I wrote about how I partnered with the CEO of Oso to fill small events by reaching out to prospects using his LinkedIn account.

What am I missing? I’d love to hear the plays that have worked for you.

A special offer for my readers

Many of these campaigns can be built and automated using Spear, a company that I am advising. I’ve seen a lot of tools in this space, and many of them try to automate the SDR job: more volume, more cold email, more junior reps. Spear is the opposite. It’s built around signals, quality, and getting your executives and AEs in front of buyers.

Spear watches for the signals behind most of the plays above: champions moving, warm paths through your team and investors, life moments worth reaching out on, things prospects said on your calls, and the people going to your events. Then it helps you act on them from your own LinkedIn, in your own voice. It’s learned from more than 5 million prospecting journeys, so it knows what a good message looks like for each moment.

Sign up through my link and you’ll get:

• Free access to Spear’s self-serve tier, so you can run your first signal-based plays from your own LinkedIn

• No hidden credit charges: you pay per profile, not per action

• A 1-hour session with one of Spear’s GTM engineers to build your first plays around your ICP

Thanks to Dor Vardi & the Spear team for contributing to this article.

I tried to follow all of the responses, but generally they fell into two categories:

1. Agree!

2. It’s capitalism, live with it

Many people on X point out that CJ owes nothing. That I shouldn’t expect integrity at the top. That employees are disposable, so why not CEOs?

Because having integrity is also good for business.

Perhaps it’s my naiveté, but I think that there is a place for integrity in business. In fact, that was something I loved about the MongoDB culture. There was an investment in developing internal talent. It wasn’t altruism, because it made the company better. (If anything, sometimes I thought we needed more external talent to infuse fresh ideas into the organization.)

MongoDB’s leadership had what I thought of as a code. It sounds old-school, but it included things like: don’t go work for a direct competitor. Have a transition plan when you take your next role, whether it’s an internal job or a departure from the company. (You weren’t getting the next big job at MongoDB without a succession plan, anyway). Don’t solicit employees when you leave the company.

That code wasn’t there to be nice; it was about building a company that both customers and employees could trust. Because people didn’t jump to competitors, we could be more open and transparent internally. Because leaders planned their exits, transitions were smoother for both customers and team members. Because departures didn’t turn into raids, people invested in their colleagues.

I’m proof of how this works. MongoDB took a chance on me by moving me from marketing to running sales operations, a bet I doubt many companies would have made. That investment made me a better and more well-rounded operator, but the company also got a return on the investment. I stayed for 15 years and brought everything I learned back into the business.

A leader’s word is a business asset. People buy from people, especially in enterprise software.

Those words are burned in my brain. It was an anonymous quote in a ten-page performance review my boss handed me right before he wrapped up his time at MongoDB. “I think you’re going to be really happy with what you see,” he’d told me.

That document included, verbatim, anonymous quotes from my peers. I could easily figure out who said what based on the context of each statement. The 154 emails exemplified the theme of the review: I could get stuff done, but I wasn’t “scaling.” I was a doer, and not a leader of people. I didn’t communicate plans, I just bombarded and overwhelmed people with my own high-speed execution.

Maybe I’m not cut out for a company that’s scaling, I thought. I’m an early-stage person. It wasn’t the first time I’d decided something like that about myself.

As an elementary school student, I excelled academically. I loved to read. Math made sense to me. I aced standardized tests. My parents and the adults around me praised me: “You’re so smart, Meghan!” they would say.

My identity as the “smart kid” crumbled in 6th grade when I sat for the entrance exam to Hunter College High School and didn’t get in. I cried all day. I guess I am not smart enough, I decided. I didn’t even want to go there, particularly; it would have meant an hour-long bus-and-subway commute from Queens into Manhattan.

Life went on. While I still mostly identified as a nerd, there was one sport I took to naturally: swimming. I joined a team at 9 years old. When I got to ninth grade, I decided to try out for the high school team, the Turtles, expecting to fit right in. The girls were lovely, but the coach wasn’t impressed with me. “Gill? Her times are terrible,” I remember him saying. After one season, I quit.

It’s only recently that I observed a pattern: not smart enough for Hunter, not fast enough for the swim team, not a strong enough leader for a scaling company. Every time, I treated a moment’s feedback about my capabilities as an unchanging thing, rather than a data point about where I currently stood.

Let’s go back to that soul-crushing performance review. I’m sure that there was some reaffirming stuff in there as well. There had to be, since my boss told me that he thought I would be happy with the feedback. A more mature Meghan would have seen this feedback the way he did (instead of as a gut punch). It was information to make me a better leader. Instead, I internalized it and took the feedback personally, thinking of it as a condemnation of my character.

Are leaders made or born?

When Dev took over as CEO of MongoDB, I was amazed by the focus on people. Recruiting and developing people were paramount, even more important than execution. With the right people in place, trained to do the right things, execution would follow naturally. At QBRs, we did talk about business performance, but spent the majority of time discussing people. Was Bob ready to be promoted, and what could we do to get him there? How could we coach Jane on this deal?

Believe it or not, this was eye-opening to me. You mean that people aren’t just born as amazing leaders?

You may have noticed a pattern, which took me years to identify and unravel. I had decided that I just wasn’t smart enough to go to Hunter, or that I wasn’t a strong enough swimmer for the high school swim team.

In one of the many training sessions I took over the years on leadership while at MongoDB, I was introduced to the concept of fixed and growth mindset. On paper it seemed rather intuitive. A person with a fixed mindset assumes that their capabilities are innate and unchangeable. The failure to compete with the high school girls implied that I just wasn’t good enough. Time to quit the swim team!

But a person with a growth mindset sees those obstacles as experiences necessary for the path to excellence.

It finally clicked for me. How many things in life have I not pursued because I wouldn’t immediately be good at them? Being a star student had drawbacks. If things didn’t come naturally, I assumed I couldn’t ever be good at them.

Over time, that fixed mindset began to unravel as I saw more evidence that people, myself included, could develop and change. I had another 360 review where my peers said I was too much of a single point of failure and didn’t have enough senior people on my team. It was reminiscent of the feedback in that ten-page report, where the team said, in so many words, that I was just a super IC and couldn’t build a strong team. Except this time, I spent the next year recruiting to prove to them, and to myself, that I could scale. I hired some amazing people on the way. (KJ, Haja, Sief, Casey, JB, Chen, to name a few.) I got promoted to SVP, recruited a successor, and gracefully departed the company after 15 years.

The growth mindset lesson that I keep learning

This whole growth mindset thing came back to bite me again recently, this time in running, of all places.

With more flexibility in my schedule in my post-MongoDB life, I really wanted to get back into shape. Before having kids I had run two marathons. My times weren’t remarkable, but my goal had been to complete the race. And I had experience swimming as a kid. Living in New Paltz, in the shadow of the Shawangunk Ridge, the SOS triathlon is the iconic fitness challenge: a 30-mile bike ride, followed by a series of alternating runs and swims through the Mohonk Preserve and Minnewaska State Park.

I spent 2025 training for the race. I didn’t expect to be competitive. I just wanted to finish, a goal that only a small number of people have achieved. I came in 11th out of 11 women in my age group, finishing in over 7 hours. “Were you in one of the later starting waves?” my mom asked as I crossed the finish line among the final finishers. “No, Mom, I’m just slow!” I responded.

I had done all the SOS training on my own, which was getting lonely. So I joined Masters Swimming at SUNY New Paltz. At the pool I met Danielle, another SOS athlete. I couldn’t help but look up Danielle’s SOS times and compare them to mine. She was faster than me on every leg of the race.

Danielle invited me on a bike ride, then a run. “I’m not fast,” I warned her. She was coming back from an injury and needed an easy pace, so I tagged along, and she introduced me to a local running group. I was the slowest one there. I’d never done real speed work in my life.

And then, something kind of amazing happened. I improved! I raced SOS again this year and finished 43 minutes faster than last year, with better times on all eight stages.

Wait a second, I thought to myself, can I also improve as a runner? I had never really thought of myself as an athlete, I was always a nerd in school. And when I was training for marathons or SOS, I didn’t focus on times, just on building the endurance to finish. Maybe it was because I didn’t want to be disappointed if I didn’t meet my goals. Maybe it was because I didn’t know how to improve until I joined a running group. But with a lot of deliberate practice, my times improved.

I had thought that I had squashed that pesky fixed mindset, but here it was, rearing its head again!

That, I think, is what all those corporate trainings on growth mindset failed to teach me. Growth mindset is not a belief you install once and keep. It’s a habit of catching yourself saying I’m not smart enough, I’m not fast enough, I’m not a leader of people, and then reminding yourself that you aren’t limited by the things you can do today. I guess it’s a lesson that I am still learning.

Then the CEO says the sales team is going to miss next quarter.

The board members are confused. Didn’t we just say the team is overachieving? What changed? Welcome to the greatest champagne problem in enterprise sales: the inability to hire and ramp sales fast enough.

Productivity may look good, but even at those productivity levels, the company doesn’t have enough sales capacity. The company is riding on a handful of fully ramped reps who were overachieving their number, and there isn’t enough of a bench behind them to carry the next quarter.

This is one of the most common, and most avoidable, mistakes I see companies make as they move past founder-led sales. Once you’re no longer relying on the founder to close every deal, forecasting recruiting becomes just as important as forecasting your revenue.

At MongoDB, we spent more time forecasting recruiting than we did forecasting revenue. By the time you’re in the current quarter, your pipeline already exists. But the reps who will carry next quarter, and the quarter after that, need to be hired and ramped now.

Building a recruiting forecast

Treat your recruiting like you treat your revenue forecast. Review it every week, the same way you’d review the pipeline. The format doesn’t need to be complicated, but it does need to be consistent. For every manager, I’d want to track the following:

• Start of quarter headcount

• New AEs starting in quarter (butts in seat, not signed)

• Anticipated not started (signed offers with a start date in the quarter)

• Attrition (expected departures)

• The sum of all of the above (your actual projected headcount by end of quarter)

Then, I’d want the sales leader’s gut call on where they’ll land at the end of the quarter. Too often, they are overly optimistic: they don’t have enough lead time to get the candidate to start within the quarter, or they forgot to account for a rep they are managing out, or any other number of things. Which brings me to my next section.

The gotchas

I think I got into more arguments about this recruiting forecast sheet than any other report, usually over tiny numbers. (Hey! I’m going to end with 5 AEs, not 4!) It’s easy for any individual manager to look like they’re “only one behind.” But once you’re at scale, with dozens of managers each “only one behind,” those onesie-twosies add up to a meaningful drop in capacity.

Here are a few patterns that I observed, both at MongoDB and at other companies that I advise.

Lack of first-line management and/or recruiting capacity

You can’t scale recruiting without frontline managers dedicated to it. I recently worked with a company that had a plan to double their AE headcount but hadn’t put the managers in place to support it first. It turned into a death spiral: the managers they did have spent all their time interviewing instead of coaching, reps didn’t get properly onboarded or developed, reps then failed and attrited… the result being that the managers had to recruit even more!

You need to lead with management so you have the capacity to recruit AEs. Ideally you can build your management bench through a mix of internal promotion and external hires. And while managers are ultimately accountable for hiring, they can’t do it without the help of a recruiting team. A productive recruiter can fill 8-10 roles per quarter, depending on the seniority of the role.

Failing to anticipate attrition

Sales is a high-turnover function.

Always assume you’ll have more attrition than you think. I recently reviewed a forecast for a company I’m working with, and they had three attrits baked in. I pushed on it. A month later, they had eight.

Build in the cushion from the start and don’t wait to be surprised. And build your recruiting capacity based on the gross hiring with a realistic assumption for attrition.

Forgetting that promos are attrition, too

You need to build your front-line management capacity, and there is no better pool of candidates than your most successful AEs. Those people are in the best position to coach new AEs, and internal promos are a huge morale booster for any team. But a promotion also means that you’re losing sales capacity.

A great way to address this: tell your promo candidate that they can get the job once they’ve hired their backfill. You’re still losing capacity, since the new AE will need to ramp up, but at least you have a butt in seat. And make sure that you also account for this in your capacity model!

Counting signed offers instead of butts in seat

I can’t tell you how many times a leader told me they were “on track” because their final rep had signed an offer but hadn’t started yet. My answer was always the same: we don’t start to ramp reps when they sign. We start ramping them when they show up for work. So if you have a May hire that signs in May but starts in June, you are late on hiring.

You always need to work backwards from notice periods. This is particularly painful in countries like Germany and India, which have long notice periods. It might be months after a signed offer that the person actually starts. Which means you need to be recruiting months in advance.

In fact, in the quarters when we had the most aggressive hiring plans, we had a “sign by date” that everyone was working towards. In the US, it was typically three weeks before the last Monday of the quarter (since all new joiners onboarded on Mondays) and even earlier for Europe and Asia. We knew that after that date, it was basically impossible to impact the recruiting in quarter, and we moved to forecasting the following quarter.

Pausing recruiting when the team is full

Even when a team is full, keep the pipeline warm and keep the operational muscle in place for managers to move forward on hires. You always need someone a few reps ahead of plan to offset the managers who are behind. If you only start recruiting once you have an open seat, you’re already behind.

Having the wrong incentives in place for managers

A manager’s own quota should be tied to their hiring plan for the year. If they miss on hiring, it’s that much harder for them to hit their number. But if they hire ahead of plan, they’re setting themselves up to overachieve. Get the incentives pointed the right way and recruiting stops being an afterthought.

What other gotchas am I missing? Feel free to comment. I’m also considering making a spreadsheet template of the recruiting forecast sheet available on Substack, so please let me know if that would be useful.

I’m hiring my second and third AEs, and I have no idea how to split up the territory. My first rep has basically had free rein of the whole market. Now I’ve got three people with three very different skill sets: one senior, closer to enterprise-ready, and a couple who are scrappier and better suited to SMB and high-growth startups.

Do I draw lines on a map? Split by industry? By company size? I keep going back and forth, and I’m worried that whatever I pick, someone ends up with a territory that “sucks,” and I spend the next six months refereeing complaints instead of running a sales team.

Signed,
Suddenly Splitting Things Three Ways

Dear Suddenly Splitting,

Congratulations! Going from one rep to three reps is a milestone. Here are a few guidelines for when you set up territories for the first time.

Territory management is the lever for productivity

Once you’ve hired the right people, territory management is your single best lever for rep productivity. Yes, this takes more upfront work: you need to research, score, and rank the best targets. But if you’re about to make a major investment in a new hire whose success is directly tied to company revenue, it’s worth setting them up to succeed. It’s far too easy for ramping reps to waste cycles chasing accounts that don’t fit your ICP; better to serve them the best accounts on a silver platter.

Give each rep a set of named accounts (not a rule)

Don’t tell a rep “you cover the Northeast” or “anything under 1,000 employees is yours.” That sounds efficient, but it just pushes the research work onto them.

Now they’re burning hours figuring out which accounts actually fit, and you’re refereeing arguments about whether a company has 998 employees or 1,004.

Focus drives productivity. Hand each rep a named set of accounts that fit your ICP. Their job is to generate pipeline against that list.

Not to mention, if you start building these territories, it can actually become a powerful recruiting tool. In an interview process, you can show candidates a list of accounts that you would assign them, and the rationale. What better way to get someone excited for the job and to ensure that they can hit the ground running once they start!

Match the accounts to the reps

While named account lists take more time to create, it also solves the issue of different rep profiles. It gives you the flexibility to apply judgement. Your senior rep who’s ready for bigger, more complex deals gets the larger enterprise accounts. Your scrappier reps get the SMB and high-growth startup accounts where speed and volume matter more than navigating a buying committee. You’re both avoiding overlap and setting each person up to win in the accounts that actually match their skill sets.

Plan for the “my territory sucks” conversation

Every rep eventually decides their list is worse than someone else’s. They will always ask for more territory. Reps assume more accounts means a bigger number. In practice, a well-designed list of 100 tight-fit accounts will outperform a sprawling list of accounts, many of which they won’t have the time to properly research and engage.

Establish a clear, consistent process for when an account can be swapped out. Set real disqualification criteria (a minimum number of attempted touches, for example) before a rep can trade in an account for a new one. And keep a “bullpen” of vetted accounts ready to go, so when someone does DQ out of a name, you’re not scrambling to backfill it.

More importantly, track activity in assigned accounts. If the rep isn’t working an account, they shouldn’t get to keep it. At MongoDB we always used to say that the CRO owns all the accounts; they are just on loan to the reps. The core message? Every day you need to prove that you are working the accounts you’ve been handed.

That combination — a curated list, sized to the team, tuned to each rep’s strengths, with a real process for refreshing it — solves more problems than people expect. It handles the “how do I avoid drawing geo lines” problem, the “how do I split enterprise from mid-market” problem, and the “how do I keep this from becoming a free-for-all” problem, all at once.

Hiring reps #2 and #3 is hard enough. Don’t let a bad territory setup undo the good hiring decisions you just made.

Rooting for your pipeline,
Meghan

To understand why, it helps to know what MongoDB looked like before there was a sales team at all. At that time, the company was called 10gen, an intentionally different name and brand from the open source software (MongoDB) that the company produced.

When I joined the company-then-known-as-10gen, MongoDB was just starting to get traction as an open source project. We had no revenue, but early adopters were kicking the tires, and distributing the software for free served as a “freemium” model that could later be monetized. I’ve previously called this the OG PLG (product-led growth) motion.

We were in the early days of the cloud, and years before we would launch MongoDB Atlas, the hosted database-as-a-service solution that is now the majority of $MDB’s revenue. The only thing we could actually sell at that time that produced recurring revenue was support: an annual subscription that gave customers direct access to 10gen engineers.

Despite the fact that we lacked a commercial product, it was clear that we needed to hire not just engineers, but also salespeople. We couldn’t keep up with the initial inbound demand, even though, ironically, we barely had anything to sell them yet. One early rep overachieved on her quota simply by getting access to a digital voicemail box we’d set up for customer inquiries, where developers left messages begging for help with MongoDB. We measured deal cycles in days and weeks instead of months. The fish were jumping into the boat.

As the sales team scaled, inbound leads grew too, just not as fast as the sales team. Instead of a few reps overwhelmed by demand and crushing their numbers, we soon had a team of mouths to feed.

Free by design

The founders’ philosophy, which I had fully internalized, was to prioritize giving developers an amazing experience. Once MongoDB had been widely adopted, then we could worry about monetization. This is a long-term-oriented and expensive approach, but the database market was and is massive – over $100 billion and growing. If we could drive enough product adoption, even monetizing a sliver of it could be hugely profitable.

We didn’t ask developers for their email address before they downloaded the software. It wasn’t uncommon to discover a well-known company running MongoDB in production without us having ever realized it.

If someone posted a question in the user forum or IRC (the precursor to Slack) at midnight, there was a decent chance someone, often one of the founders, would answer it by morning.

We wanted MongoDB to feel like a community-run open source project, and not at all corporate. 10gen and MongoDB were different brands, each with their own websites and color palette.

The net result was that developers could download MongoDB without even leaving their name, they could get answers to most of their questions for free, and they weren’t even aware that there was a company called 10gen behind the software selling support.

Even if they did know about 10gen, many early adopters didn’t see the point of buying support. The sales reps weren’t wrong when they complained that we were our own biggest competition. Why pay for support when you could get it for free?

The culture mismatch between product and sales

As far as I could tell, 10gen’s founder, Dwight, wasn’t losing sleep over any of this. For him, widespread adoption was the goal, and monetization was a problem for future-Dwight. But the reps had quotas today, creating a very real culture mismatch. The company had been built by engineers, run by engineers, and optimized for engineers. I had barely felt accepted as a marketer on this highly technical team. Sales people were a different animal entirely, and they had opinions about everything:

• Sales wanted to gate the download (i.e., require a person to share their email before downloading the software) so that they could more easily prospect. PLG says remove friction. Sales says friction is how you find out who your prospects are!

• Sales wanted engineering, and the founders in particular, to stop providing such great support to open source users for free. PLG says the support is the product experience. Sales says free support is a subscription you’re giving away.

• Sales wanted differentiation between the open source product (which anyone could get for free), and the product that customers could buy when they purchased a subscription. Support simply wasn’t differentiated enough. But every feature we kept free was also a feature that drove adoption, built goodwill, and made the community bigger and stickier for the long term.

• Sales wanted to get rid of the 10gen branding so that they could call on prospects as “Valerie from MongoDB” instead of “Valerie from 10gen.” To sales, having two names for one thing created friction and confusion. But from a PLG standpoint, keeping the open source and commercial entities separate reinforced MongoDB as a developer-focused, community project.

Here’s a tangible example of that friction. In a flash of his typical creativity, Dwight had me add a form to the website offering “lightning consults” — one-hour sessions with a 10gen engineer for $400. It was a clever way for us to make sure that any company using MongoDB could quickly get help. Ironically, even though this generated a small amount of revenue, sales hated these even more than the free support. Instead of getting a customer onto an annual subscription, we were now undercutting the sales team, selling support one issue at a time!

Looking back, almost every one of these exemplified the inherent tension between the PLG instincts (frictionless access, community trust) and sales (visible pipeline, hitting quarterly targets).

There’s a fine line between clever and stupid

In the 1984 mockumentary Spinal Tap, the band famously says there’s a fine line between clever and stupid. We spent years walking that line. How much information could we gather about the developers using our product without making them feel exploited? This is really the PLG-versus-sales tension in miniature: every tactic below was an attempt to give sales what it needed without breaking the trust that made developers want to use MongoDB in the first place.

For example, instead of gating the download behind a form, we put a form after the download, asking if people wanted to stay in touch. That generated hundreds of sign-ups a week, and I set up a drip of emails educating those developers on MongoDB and inviting them to reach out if they needed help. In 2013 we launched MongoDB University, offering free online courses for developers and DBAs. (Last I checked, over a million people had signed up for one of these courses.)

Eventually we started using tools to deanonymize our web traffic. We weren’t trying to identify individual people, but the companies behind the downloads so we could arm sales with that information. A rep could go to a CTO and say, Did you know your team has downloaded MongoDB fifty times this month? We should talk.

We also drew a line inside the product itself: developer-facing functionality would stay free and open source; the security and operations features that ops teams needed to run MongoDB at scale would be commercial. That distinction became the seed of what eventually grew into MongoDB Enterprise Advanced.

These were all sales trying to retrofit itself onto a PLG motion. As long as the best version of the product was free to download and support was easy to access, the tension wasn’t going away.

Chasing the smoke (as in, where there is smoke, there’s fire)

All the clever micro-tactics, like semi-gated forms, deanonymizing traffic, free online courses, were ways of trying to monetize a fundamentally free thing. What finally worked? It was Atlas, the hosted database service launched in 2016.

Before Atlas, our sales team had mostly chased large, regulated industries who’d pay for commercial licensing and enterprise support. We were typically selling to the operations persona, folks who are generally risk-averse and hence resistant to change. And smaller companies rarely saw the value in support; they’d just run the database themselves.

Atlas changed who we were selling to. We went from selling to ops teams who had to be convinced to buy an insurance policy (i.e. support), to selling directly to developers who wanted someone else to manage the database for them. Enterprises were still in the mix, but the aperture had opened to startups and digital-native companies that had never been a real prospect before.

Our lead database was full of people from companies that had downloaded the open source product and been disqualified by sales because they wanted to DIY. They were self-managing MongoDB on their own infrastructure instead of paying for support. Overnight, that list of rejects became the most valuable list we had. Every one of those self-managed deployments was a company already sold on MongoDB but still doing the undifferentiated heavy lifting of running it themselves, which was exactly the pain Atlas was built to take off their hands. Sales didn’t have to convince them MongoDB was the right database. They just had to convince them to stop managing it themselves.

None of this demand appeared out of nowhere the day Atlas launched. It was the payoff of years of building an open source community. By this point, I had transitioned to running sales ops, where my team was tasked with building “smoke reports.” These were lists of companies actively engaging with MongoDB, pulled from download activity, community touchpoints, job listings, and other third party data. Where there’s smoke, there’s fire, we trained the salesforce. We used those reports to build sales territories. We weren’t cold-calling; we now had something valuable to sell to these organizations who had adopted MongoDB.

This is where the strength of MongoDB’s sales culture shined. Armed with this data, sales was disciplined in generating pipeline against these warm accounts. It was a brilliant formula for a successful go-to-market. Sales didn’t need to convince prospects to use MongoDB, they already knew and loved the product. Their job was to get those MongoDB fans to form a commercial partnership with us.

So… can sales and PLG co-exist?

The sales machine got very good, very fast. Maybe too good. In my final years at MongoDB, Postgres started closing the gap as one of the most popular databases, becoming the default database for a lot of startup developers. We still had years of latent demand left to convert, so Postgres’s rise was slow and steady; easy to miss until it wasn’t.

A PLG motion generates enormous value that doesn’t show up immediately as revenue, and I’ve often wondered whether we’d gotten so disciplined at converting demand into sales territories and quota that we stopped tending to the thing that created the demand in the first place: the years of community work, free support, and goodwill that gave us the “smoke signals” in the first place.

We got very good at harvesting the field. But had we forgotten that we needed to invest just as much in planting? This is the conundrum of investing in community and developer relations at many companies. It’s expensive and very difficult to measure ROI, with the impact measured in years instead of quarters.

Product-led growth and sales, by nature, are optimizing for different things. Early on, we arguably weren’t commercial enough — a frictionless product and years of free support, with no real machine behind it to convert any of that goodwill into revenue. By the end, we may have swung too far the other way.

I don’t have a tidy conclusion here because I don’t know that one exists. I don’t think that there is a perfect formula for building a company that’s both product-led and good at sales. It’s a needle you keep threading, for as long as the company exists.

Me? I was on the marketing team. I never would have considered running RevOps as an option, but to his credit, MongoDB’s former CEO (Dev Ittycheria) was all about developing people. And that often included lateral moves.

“We need a new head of sales operations. I think you would be a good candidate,” Dev told me in early 2017. I was surprised. I vaguely understood the function from working with the Salesforce admin team to track marketing leads through to closed business. And because I owned demand generation, I spent a lot of time looking at data. But that was the extent of my qualifications.

I got 30 minutes with our very busy CRO the next time he came to New York. It was a bit awkward. I was intrigued by the idea of trying something new but also felt out of my depth. And there was a complicating factor I wasn’t sharing with anyone yet: I had just found out I was pregnant. It was too early to tell my parents, let alone the very-busy-CRO currently sizing me up for a leadership role. Conflicted, I decided to go through the motions anyway.

The previous head of sales ops was a former investment banker, who used to joke with me that he liked to work half days: from 7am to 7pm. Mr. 7-to-7 had just left MongoDB to take on a CEO role, and I called him to get his advice on taking the job.

“I have two questions for you,” Mr. 7-to-7 said. “First, do you like Excel?”

“Sure,” I said. Excel and I had always been friends. I’d spent much of my time in marketing analyzing lead cohorts and conversion rates.

“Good. You’ll be in Excel all the time.” Then: “Second, are you prepared to spend 50% of your time dealing with sales compensation?”

“Sales compensation? How complicated could that be?”

He just laughed at my naiveté.

The Alien in the Room

I joined MongoDB as the first “non-technical hire.” That’s what they called me, anyways. I felt like an outsider among a team of low-level nerds. But if I’d felt like an outsider among the engineers in 2009, years later I felt like an alien in the sales organization. The entire leadership team was male, as were nearly all the individual sales people. Everyone talked openly about money and their commission checks. And where I used to feel like an imposter for never having written code, I now felt like an imposter for never having “carried a bag.”

We were coming off a weak quarter. Morale was low, and several sales people had exited the company. With fewer sales reps, and hence less sales “firepower,” the sales leaders were worrying about how they would hit their numbers.

My first assignment was to placate the sales leaders. “They need new quotas,” the very-busy-CRO demanded. “We have to adjust their targets, or we are going to lose even more people. And we’ve hired new sales managers who don’t have comp plans yet. They’re getting restless. Can you handle that?”

What exactly is a comp plan, I wondered. I got access to a shared drive full of compensation files: one Excel spreadsheet per sales leader, each with a row for every rep or hire on their team, and a quota calculation dense with nested IF-statements. I opened the first one and stared at it for hours, trying to reverse-engineer the logic.

I guess this was why Mr 7-to-7 had laughed when I asked how complicated sales compensation could be.

By 11:30 pm that night, I put my head in my hands at my tiny New York City dining room table. A pit formed in my stomach. I had taken the job without realizing how little I knew, and I had to learn it before I had a baby. Going back to my old job in marketing wasn’t an option: the CMO was already close to hiring my backfill. I’ve made a horrible mistake, I thought.

“I don’t think I’m cut out for this,” I told my husband.

Pregnant and with parental leave looming, I didn’t have a choice. I sat at the dining room table with pen and paper, walking through each cell and scribbling flow charts to represent the branching logic.

Colorblind

The start of each quarter was the busiest stretch for my RevOps team. We would convene in person to tick and tie every last attainment metric and build the deck recognizing the top performers for the Sales All Hands. The very-busy-CRO insisted this presentation happen the Monday after the quarter closed, no exceptions, which often meant working through the weekend.

It sounds simple: pull the numbers, highlight the top performers, present the results. It was not simple. Did we count ramping reps or only fully ramped ones? Did we properly account for deals split across teams? Had finance adjusted the terms on a given opportunity before we pulled it? There were a hundred reasons why you couldn’t just export a table from Salesforce and call it done.

I dreaded those All Hands calls. It was far too easy for a slide to say $356,000 when the correct number was $365,000, or for someone to be furious that we’d ranked the top reps by percentage-to-quota instead of absolute dollars. For a job that was supposed to be purely data-driven and analytical, it turned out that salespeople have a lot of opinions about data.

After the All Hands, sales managers ran Quarterly Business Reviews (QBRs) with their own reps. By week three of the quarter, the very-busy-CRO ran QBRs with the sales leaders. We’d pull up something called the “Red Yellow Green” — a chart color-coding every rep by whether they’d hit their number across the last five quarters, and the very-busy-CRO would ask each manager who was ready for promotion, who needed development, and who needed to be managed out.

My team produced all the metrics: attainment, pipeline, the Red Yellow Green, and a dozen others. The Red Yellow Green in particular was a source of constant low-grade panic for me. It was mostly automated, but the automation had blind spots: a rep promoted into a new role and territory, for instance, would get flagged red by a formula that only knew start dates, when in reality they were ramping into the role or territory. The very-busy-CRO wasn’t shy about pointing it out when the colors were wrong.

After the second round of QBRs, he pulled me aside, and he wasn’t happy. “The data quality isn’t good enough,” he said. After sitting in on so many QBRs where we evaluated people’s performance, I could read between the lines: get my shit together or we find someone else for the role.

I relayed the feedback to James, the sole analyst on my team, who built the charts. “We’ve got to do a better job with the data,” I told him. “We’re making too many mistakes.”

That’s when James mentioned, almost in passing, that he was colorblind.

“Well,” I laughed, “that explains why QA-ing a chart called Red Yellow Green might be a challenge for you.”

I tried to memorize the various edge cases that could throw off the data. And still, the very-busy-CRO would call something out mid-session and turn to me. I’d start frantically digging through spreadsheets and Salesforce reports while he grilled the regional sales director standing in front of me.

I was never on stage the way the sales leaders were, but I felt evaluated in every QBR. The days ran long, and I never knew when someone would turn and ask me for a number on the spot.

Always Be Planning

In my first several months on the job, compensation exceptions were landing on my desk constantly. Our salaries weren’t at market. The very-busy-CRO and his team wanted me to bring forth a proposal for a market adjustment as part of our annual planning process.

Dev was not pleased with my proposal. What are you talking about, he said, while, I later learned, he was texting the CEO of a peer company to compare salary bands. Making this change would cost a lot of money. Where’s the data showing this is actually necessary?

I left that meeting embarrassed. It was among the first of many annual planning tasks that kept getting longer: segmentation of the market, territory carving, quota modeling, headcount requests… all of it stacked on top of a quarterly rhythm I still hadn’t fully caught up to.

Afterwards, I went to the recruiting team, and together we pulled every offer we’d issued in the past year and how many had needed non-standard compensation to close. Recruiting built out a comparison set of companies and their salaries. I rebuilt the proposal and brought it back.

I was still commuting on the subway to the office, climbing up about 50 steps to reach the elevated train at the Astoria Ditmars Blvd stop. I came into the office for a planning session two days overdue on January 4th, 2018. Scrolling through my old photos I found this picture of Dev in front of a white board, writing about who-knows-what:

At the end of that session, I told my colleagues that I would “see them tomorrow, maybe.” And then went into labor that night, giving birth to a beautiful baby girl.

If I thought I was unprepared to run RevOps, I was about to find out what real unpreparedness felt like. There is no ramp time with a newborn. It turns out qualification didn’t come first in either job. I got handed the thing, and the readiness came after (if it ever did).

I stuck with the RevOps job for seven more planning cycles. And I now have three kids, ages 8, 6, and 4 years old. Last week my cousin had his first child.

Thanks to “James the analyst” for reviewing an early version of this article. He is now running ops for Profound, an AI marketing intelligence platform, where he probably had an analyst use Claude Code to build the RYG in 30 seconds. And of course to my better half Graham for reviewing.

How to approach your first sales conversations as a founder, the difference between outbound and inbound, building a simple repeatable sales process, running a discovery call, handling objections, knowing when to hire your first sales person, and transitioning from founder-led sales to a scalable motion.

That’s a lot of ground to cover, so I tried to distill it into a few simple themes.

Work your network before you work a list

My first real piece of advice was the least glamorous: do things that don’t scale. In the earliest days, founder-led sales is personal outreach and manual onboarding.

Before you touch a cold list, exhaust your warm network. Someone needs to take a chance on you, and a personal connection makes it way more likely that they will. Professors, classmates, accelerator peers, advisors, speakers who’ve come through the program are all options for warm intros.

And for the love of God, don’t spray and pray, especially not when you are trying to source design partners and understand the customer. Until you really understand the buyer, it doesn’t make any sense to try to scale up outreach.

Form a hypothesis about your ICP

Once you’ve mined your network, you need a hypothesis to begin scaling outbound against: an Ideal Customer Profile (ICP). Something like “SaaS companies with a microservices architecture, US-based, targeting platform engineers,” which is specific enough that you start to learn the language your buyers use, where the pain actually lives, and where they spend their time.

Some founders get wrapped around the axel because their ICP might be much more narrow than their overall TAM. That’s ok, the ICP is a starting point, a place where you can start to pattern match and build expertise.

From there, expect to make at least 9 or 10 touches: A personalized LinkedIn note, a warm intro through a mutual connection, a cold email, a phone call, a voicemail that reinforces the other touches.

Discovery: you’re there to learn, not pitch

Once you’re on the call, the job is to listen more than you talk. Your goal is to understand the prospect’s pain. How are they currently solving the problem? What happens if they continue as is? What would the impact be on making a change - to revenue, to the budget, or to employee productivity.

But most importantly is to find the pain. No pain, no deal.

I gave an example when I was running Sales Ops at MongoDB. I did a demo of a conversational intelligence tool (I won’t mention which one). They had killer features and a better user experience than the solution we had in place. So the rep was shocked when I told her I didn’t want to move forward.

“This isn’t a pain point,” I explained. “Your solution might be superior to the one in place, but there are other problems that are much more urgent.”

You should leave the discovery stage of the deal with the able to answer the “Three Whys”: why anything (is there real pain here worth solving), why now (what’s the urgency), and why us (what’s the differentiation).

Record and transcribe every call if you can. That data is gold, and useful for iterating on both your sales messaging and marketing copy.

Champion vs. coach

What’s the most important sales skill? Is it the ability to pitch? To negotiate? Or simply the ability to connect with people? No. Qualification is the most important skill in sales.

Founders can waste a lot of cycles with “happy ears.” Your time is your most valuable asset, so make sure that you are prioritizing prospects with real pain. One great call is not a deal.

We discussed the distinction between a coach vs. a champion. A coach is friendly, gives you information, helps you understand the org chart, but has no power to move the deal. A champion has power and influence and has something personal riding on the deal’s success.

No pain, no champion, no deal.

Be paranoid about how they actually buy

Ask your champion to walk you through the procurement process. Then go looking for reasons the deal won’t happen, instead of waiting for them to surface on their own: security review, procurement, legal, final budget sign-off.

I learned from sales legend John McMahon that you should never lose a proof of concept. Influence the criteria before you enter one. Agree in writing on what success looks like (”if we can do X, Y, and Z in the next six weeks, then we agree that...”) and run parallel tracks (security, legal, procurement) during the POC.

Hiring your first sales rep

The rule of thumb I gave: you should be able to sell something like $1M a year yourself, with some real consistency in how you’re closing, before you hire. The Founder Quota Rule is a good gut-check if you want a more formal version of that.

And when you do hire, don’t expect to hand sales off. Your first account executive needs your pipeline, your relationships, and your coaching to have a shot at succeeding.

Thanks for having me NYU! You can also check out their recap on the Entrepreneurial Institute blog.

The case for the marketing engineer:

• Using AI across the marketing stack is a specialized skill set

• An AI-native marketing approach requires someone a direct reporting line to the CMO or VP of Marketing who can work cross-functionally

I’m not convinced that this is a net new job, at least not in the long-term, so here’s my hot take.

AI changes every marketing function

I do expect that marketing teams will be required to have more technical skills.

Product marketers will use AI to run competitive research and build battlecards in an afternoon instead of a month. Content marketers will use it to iterate, repurpose, and atomize a single asset into ten formats. And so forth, across analytics, design, communications, and all of the other marketing sub-disciplines.

The AI skills get absorbed, the same way “knows how to use a computer” stopped being a job description and became table stakes.

Cross-functional ICs are how you create confusion, not clarity

We’re not in the ZIRP era anymore, when you could throw headcount at every emerging need and sort out the org chart later. That era created some bad habits. I once worked with an organization that had a dedicated operations team, a chief of staff, and a center of excellence — all staffed with talented, well-meaning people, all stepping on each other. Instead of accelerating output, this structure led to diffusion of responsibility.

A “marketing engineer” layered on top of product marketing, content, and demand gen has the same failure mode baked in.

Marketing already has a technical function

High-functioning marketing teams already have a technical function. It’s called marketing operations. If the argument is “marketing needs someone who can build systems, automate workflows, and think in pipelines,” that role exists and has for years. We don’t need to invent a new title to describe it, or to duplicate many of its core functions with a new part of the org chart.

AI has failed to to kill the SDR

We just lived through a near-identical cycle with sales development. A couple of years ago, the consensus was that AI was about to make the SDR extinct. The “AI SDR” became its own venture category. By some estimates the market has grown into the billions, with a wave of startups promising to replace the human top-of-funnel entirely.

Yet, a quick job search will find AI-native companies like Clay, OpenAI, and Anthropic actively hiring for human SDRs.

6sense’s State of the SDR research found that SDR teams using AI were actually growing and performing better, not disappearing. Not to mention, one of those AI SDR companies was found to have fabricated its metrics in a highly-publicized scandal.

Matt Heinz captured this in his piece I Was Told BDRs Would Be Extinct By Now. Someone told him years ago the role would be dead by now; it isn’t, but the actual work has changed enormously. That’s exactly right, and it’s the template for how I think about every “AI is creating/killing a role” prediction.

The same logic applies to the GTM engineer, the other title everyone’s excited about right now. It tends to live at the intersection of SDR and ops, using AI to research and target prospects. Useful work! But again — is that a new species, or is it what every AI-native sales rep or SDR will look like in the future?

So, are they real jobs?

For now, yes, but in the same way “growth hacker” was a real job in 2013. But I’d bet against “marketing engineer” and “GTM engineer” surviving as distinct roles five years out. The skill won’t stay concentrated in one headcount line; it’ll diffuse into the job description of everyone around it.

When I joined MongoDB as the first go-to-market employee, I worked directly with the founder, Dwight Merriman. Dwight didn’t hand marketing off when he hired me; he understood developers and the best way to reach them. He asked me to organize a developer event in San Francisco about MongoDB, which eventually launched a series of MongoDB events. He thought coffee mugs were the best kind of swag; I eventually ended up mailing them all over the world to MongoDB community members. He would tweak the language on the homepage himself, working through whether MongoDB should be described as a “schema-free” database, a “flexible schema” database, or a “dynamic schema” database.

I’ve spent a lot of years since then on the other side of this conversation, talking to founders about when and how to make their first marketing hire. The failure modes are pretty consistent, and last week, Justin Dunham and I led a session for founders at Base10’s San Francisco on the topic. Here are the failure modes that we discussed:

1. Hiring a marketer to solve a product problem

If your product doesn’t clearly solve a real problem, marketing will struggle regardless of who you hire or when. Good marketing accelerates a product that’s working. It rarely rescues one that isn’t.

I know this sounds obvious. But a surprising number of founders hire their first marketer when the problem they’re actually trying to solve is: why isn’t anyone buying this? That’s a product question. A marketer can help you talk about what you’ve built, but they can’t fix what you’ve built.

For most B2B companies, wait until you have some closed business and a meaningful sense of who’s buying and why. You don’t need a sales team in place first, but you do need signals that something is working.

2. Hiring the wrong profile

Marketing is a function with many distinct sub-disciplines: product marketing, demand generation, marketing ops, communications, brand. These are not interchangeable.

Hiring a great demand gen marketer and expecting them to own your positioning is like hiring a radiologist and asking them to staff the emergency room. They’re both doctors. One diagnoses, one treats. One might do okay in place of the other in a pinch, but your expectations should be very different depending on who’s actually in the role.

For most early-stage B2B companies, the decision comes down to two profiles:

Product marketing is about what you say and to whom: positioning, messaging, website copy, sales enablement, competitive framing. The output is a clear, differentiated story the rest of the company can actually use.

Demand generation is about getting people to hear it: content, SEO, paid, events, the channels and mechanics of pipeline creation.

The right call depends on what’s actually blocking you. If everyone who hears about you loves it but almost nobody has heard of you, hire demand gen. If people are hearing about you but leaving confused, you have a messaging problem, hire a product marketer.

The failure mode is hiring “a marketer” without thinking this through, and then being disappointed when the person you hired isn’t solving the problem you needed solved.

3. Hiring a VP when you need an IC

Even once you’ve figured out the right profile, many founders default to hiring a VP.

A VP-level hire at this stage typically wants to build a team and delegate before you’re ready. They’ll write strategy documents when what you need is someone who writes the brief, makes the creative call, and ships the thing. Not to mention, most VPs of Marketing bias towards one of the marketing sub-disciplines above. So if you hire a VP and hire for the wrong skill set, you will have 2x the problems.

The right first hire is a strong individual contributor with enough experience to own strategy and enough drive to execute it personally. With modern AI tools, one senior marketer can produce what previously required a team, including writing, research, analytics, basic design,and basic reporting.

4. Getting excited about logos on a resume

A lot of founders get excited by recognizable names on a LinkedIn profile. Someone spent three years at Salesforce or Stripe and they get excited.

But did they drive that company’s success, or were they simply along for the ride? Large marketing orgs have room to specialize in ways that small ones don’t. Someone who “owned demand gen” at a 300-person company might have managed vendors, attended standups, and reported on metrics someone else set. That’s a very different thing from building pipeline from zero.

5. Delegating positioning instead of partnering on it

Founders are closest to the product, the customers, and the problem. A good product marketer can get the story out of your head and into a format the rest of the company can actually use — website copy, sales decks, competitive battlecards. What they can’t do is generate the insight from scratch. They need ongoing access to you, to customers, and to the feedback coming back from sales.

At the start of this article, I mentioned that Dwight used to iterate on the words we used to describe MongoDB. Another great example is Charity Majors of Honeycomb. She didn’t just build a product; she spent years developing the concept of observability, drawing on how the term was being used in hardware and working out what it meant for software systems. In a recent conversation on messaging, Charity commented that “as soon as something starts working, every other company rushes to copy you. It’s a moving target.” She intuitively understood that messaging is a never-ending challenge.

6. Using AI to reinvent the wheel

A question came up at the event that I’ve been thinking about since: a founder mentioned they weren’t publishing content fast enough. They were running into friction with their CMS and felt like they should build something custom.

AI is useful for reducing toil. It is not a good substitute for off-the-shelf tools that already solve hard problems well. Using AI to rebuild infrastructure that already exists is a great way to waste engineering cycles. Webflow, Contentful, WordPress — these products work just fine for a lot of marketing teams, and yours probably isn’t an exception. If publishing is too slow, the right use of AI is to automate the parts of the process that are actually slowing you down. Build an agent that handles the last-mile tasks: formatting, tagging, scheduling, cross-posting. Let the marketer focus on writing and editorial judgment and automate the mechanical steps around it.

7. Skipping the work product in the interview

This is true of any hire, but I feel strongly about it for marketing specifically: ask candidates to produce something during the interview process.

It’s good for both sides. You see their actual work, not their ability to talk about work. They learn what the job will actually feel like. If someone is wrong for the role, a work sample surfaces it faster than references.

My standing example: when I was at MongoDB, I was hiring for a demand gen coordinator role. Candidates had to take an existing white paper and turn it into several assets, including a blog post, an email series, and social media posts. One candidate’s writing popped off the page, even though he didn’t have a marketing background at all; he’d been a technical recruiter. I would have missed him entirely under a standard interview process. He’s now VP of Revenue Marketing at Astronomer.

8. Overcomplicating the analytics stack

Early-stage marketing teams don’t need sophisticated analytics infrastructure. They need enough signal to make decisions.

The starting stack can be pretty simple: GA4, Google Search Console, Bing Webmaster Tools (the only source right now for data on what’s being cited by LLMs), and Hubspot.

In the early days, you’re looking for signals, not proof. A few worth tracking:

Branded search volume in Google Search Console: when more people search your company name organically, your market presence is growing. It’s a leading indicator.

LLM visibility: check whether you appear in ChatGPT, Perplexity, or Gemini when someone searches for topics you care about. If you’re not showing up there, you’re on shortlists that will never appear in your analytics.

Self-reported attribution: a simple open-text field on your demo form (”How did you hear about us?”) captures things attribution software never will.

Content connection to pipeline: build a report in HubSpot that shows which content signups or demo requests tend to engage with the most. This tells you what’s actually influencing buyers rather than what’s just getting traffic.

That’s a wrap! Thanks to Ansa (where I am a Venture Partner) and Base10, who co-hosted the event that this blog post was based on. More on this topic:

• When (and how) should I make my first marketing hire?

• How do I know if marketing is working?

• Failure modes for your first marketing hire

• Why the First Marketing Hire Usually Fails

First takeaway: recent articles with superlatives are more likely to get cited. Hence the title of this post :-P

Kidding aside, The State of AI Search from Profound’s research team was jam-packed with charts and stats. Here are some of the interesting data points from that session.

Ads have arrived in ChatGPT, and the inventory is tiny

• Only 30% of eligible ChatGPT users currently see ads. 47% of ad placements land on the first turn of a conversation, and 83% of conversations carry only one ad placement. While Instagram is serving roughly one ad per minute, ChatGPT is serving one ad per conversation. Scarce inventory means CPMs are running about 4x Meta’s.
  

• The broke out conversations into buckets: generative (43%), informational (36%), and commercial (20%). All three generate ads, but commercial prompts trigger ads 1.5x more often.
  

• The biggest determinant of ad relevance is your title and description, beating out even your landing page. Copywriting matters!
  

• Unrelated to ads: In May, ChatGPT referral traffic jumped 60% when citations became clickable links. A quarter of those clicks now land on homepages (up from 3.6%). Your homepage matters!

Claude is criminally underrated (their words, but the data backs it up)

• Ramp’s spend data shows more businesses paying for Anthropic than OpenAI, which opens an opportunity for marketers who can understand how Claude works.
  

• Unlike ChatGPT, Claude leans heavily on traditional search signals. The result is that the overlap between what ChatGPT and Claude cite is only 8% at the domain level. (Said another way, 92% of what ChatGPT cites, Claude doesn’t.) Overlap between Claude and Google is 64%. (Makes sense, given that they are the search engine giant.)
  

• One simple takeaway: being ranked on Brave is how you influence Claude. 79% of Claude’s citations came from Brave’s top 10 search results. You can run searches in Brave for free to see where you stand.
  

• Claude only searches the web about a third of the time. You can only optimize for Claude when it actually searches, so target the prompt types that trigger search, such as…

  

• One detail that can make an immediate impact: Claude stamps its fan-out queries with years. Put the year in your title and you’ll match more naturally. (See: the title of this post.)
  

• This slide summarizes how Claude and ChatGPT prioritize sources for citations:

  

Reverse-engineering Google’s AI Mode

• Google moved AI Mode into the main search bar, so this is now mainstream behavior.
  

• Google’s approach is binary: for some queries it reads its own data, for others it reads third-party sources. “Best startups” looked at the open web; “best shoe store” looked at Google’s own data (customer reviews and the like).
  

• Be specific in your copy, because Google tends to look at page summaries, and you don’t want details like the product name or location to get lost. Here is the example they shared:

  

• You can reverse engineer AI mode. Pull up AI Mode responses for your key prompts and read the headings. Those headings are the winning themes for that topic. (The example they gave was shoe stores, where the response was broken out into independent stories, specialty stores, and large chains.) Then build content to those themes.
  

This area is moving fast, so I’ll make sure to keep writing new articles on AEO with superlatives in the title!

Thanks to James Underhill for getting me a seat at this sold out conference!

Here is another example, this time espousing an agentic AI platform to help grow pipeline. The irony is that it’s clear that the sender only personalized by putting in tokens for my name and company:

Savvy sales reps can now use tools like Clay to download massive lead lists and blast out thousands of “personalized” emails that reference your LinkedIn headline, as if that counts as research. It seems that instead of making sales more targeted, we are just filling everyone’s inbox with even more crap.

The targets of these automated emails are getting more savvy, too. There have been some hilarious examples of people putting prompt injections into their LinkedIn profiles to cause automated recruiter emails to address them in Old English or send them recipes for flan.

From Casting a Wide Net to Fishing With a Spear

Every piece of content on B2B sales and marketing best practices talks about messaging “the right person, at the right time, with the right message,” yet very few companies accomplish this. The dominant approach has gone in exactly the wrong direction with the proliferation of AI tools. A single rep can now blanket thousands of prospects in the time it used to take to compose a single precise message.

That’s what drew me to Spear. Where most tools ask “how many can we reach?” Spear asks “who actually deserves a message right now, and why?”

I’ve been experimenting with Spear for months now, specifically in the context of founder-led sales at Oso. (For the uninitiated, Oso is a Series A, Sequoia-backed startup funded by my better half, Graham. Oso provides visibility and control into agents.) Spear is built on the premise that a few well-timed, relevant messages to the right people are worth more than a thousand generic ones.

How Spear Works

Spear connects to the founder’s (or another executive’s) LinkedIn. This matters because when you are trying to reach senior buyers, they are far more likely to respond to a c-suite executive than an account manager or sales development rep.

Spear learns your ICP and personas. When you set up Spear, you give it some basic information about your ideal customer profile and personas. Spear will then make daily suggestions of people to connect with, which you review. As you approve or reject suggestions, Spear gets better at figuring out the right people to suggest.

This review step isn’t just something Graham and I chose to do. It’s a core part of how Spear is designed to work. The product intentionally keeps a human in the loop, using your feedback to calibrate and improve over time.

Every outreach is tied to a compelling event. Rather than “I noticed you work in [industry], wanted to connect,” Spear surfaces a relevant reason to reach out: a new role, a recent news announcement, attendance at an event, etc.

Using Spear for Exec Dinners

The use case where Spear has really proven itself is executive dinners.

Having been part of both sales and marketing organizations, I’ve seen the same story play out many times. Sales and Marketing agree to organize a pipeline generation dinner. The hardest part of organizing these kinds of events is filling the table with the right people. Sales expects marketing to fill the room. Marketing expects sales to fill the room. A few days before the event, the teams are panicking because they don’t have enough people RSVPed.

At Oso, we built a playbook where we were filling dinners on a monthly basis with high-quality attendees, without a sales team.

We’d start by picking someone, such as an executive at a customer account or a friendly investor, to be the main attraction. Then we’d run a Spear campaign targeting people who’d be a good fit for the dinner: same ICP, right seniority level, geographically close to the event.

Every day, Spear would recommend a small but highly tailored list of prospects. Graham and I would review together, and decide which ones we would connect with and with which message. (Over time, once Graham and I were well calibrated on the prospects and messaging, I did this on his behalf every day.)

The outreach wasn’t “come hear our pitch.” An example message might look like the below, and we had the flexibility to customize each message:

“Hi [Name], I’m reaching out to invite you to a breakfast we are hosting for engineering leaders. [Famous engineer leader] who [did something exceptional] will be there. I think you’d enjoy the convo. Happy to connect either way.

If the person accepted the connection request, we had an automated follow up shortly thereafter with more details on the dinner.

We were successfully filling dinners regularly with great prospects, and there was even an unanticipated benefit: running campaigns through Graham’s LinkedIn profile grew his network of relevant executives. Graham was already writing a lot about the domain and publishing on LinkedIn. Now that thought leadership would reach a larger audience.

The Takeaway

AI is good at pattern recognition: finding the right profile, surfacing the compelling event, drafting the message. But the judgment of whether this specific person deserves your attention right now, and whether you can say something relevant to them, still requires a human.

Spear isn’t a “set it and forget it” solution. Graham and I reviewed suggestions together every day. The AI handles the research and sequencing; the humans decide who’s worth reaching and why.

Even with the benefit of AI, outbound is still hard. The reps who win won’t be the ones who automate the most. They’ll be the ones who know when not to.

TL;DR

• Moving to usage-based pricing doesn’t just change how you bill. It breaks several foundational operational processes, particularly pipeline tracking and forecasting.

• Pipeline is the first thing to break. Traditional opportunity objects don’t work for consumption models.

• ACV-based forecasting with weighted stages becomes irrelevant. Forecasting requires a new model: more analytical, less deal-by-deal judgment.

1 — Pipeline Tracking

How do you track pipeline in a consumption-based sales force? This is one of the most common questions I get from CROs and heads of Sales Ops. In the traditional SaaS model, reps create opportunities with SKUs, prices, and close dates. That makes it easy to understand how much pipeline reps are creating (the leading indicator of bookings) and how much is closing in the quarter (your pipeline coverage against your forecast).

In a usage-based model, those metrics don’t work. A rep may have a customer who already has a contract and is using the product. The rep’s job is to get that customer to increase usage by adding new use cases or features. That requires a sales cycle, but it doesn’t produce a traditional opportunity with a dollar value and a close date.

In companies at the start of their usage-based journey, I’ve consistently seen the same workaround: reps create dummy contract opportunities to track their sales cycles and show their managers they’re creating pipeline. The problem is that contract opportunities follow a specific sales process and are expected to close with a signed contract. When reps use them to track incremental usage instead there is no signed contract. Eventually reps mark them as closed lost, creating CRM clutter and dirty data.

The fix I’ve seen at least three companies land on is a dedicated opportunity type specifically for tracking incremental usage. At MongoDB, we called these New Workload opportunities. Other companies I’ve spoken with call them “Use Cases.” The key characteristics:

• No SKUs, no finance sign-off — these are internal tracking objects, not contracts

• This opportunity captures the expected incremental ARR, expected spend start date, and other key data points

• The opportunity closes when the sales cycle is complete and the use case is handed to the post-sales team (typically the CSM)

This gives you something to track. But it doesn’t fully solve the problem, which brings me to the next challenge: attribution.

2 — Attribution

Let’s say you set up the custom opportunity type described above. What happens after that opportunity is handed off to the post-sales team? How do you know if it actually generated incremental spend?

For a new logo, this is easy — you can attribute all growth in that account to the initial sales cycle. A cross-sell with a distinct SKU is also straightforward. Where it’s tricky is when you have multiple use cases within the same customer account, all running through the same meter.

The ideal state is having attribution built into the product that ties incremental consumption back to the originating opportunity. But products are rarely designed with sales attribution tracking in mind — and you could argue they shouldn’t be, since R&D time is better spent on customer-facing functionality.

The practical result: the standard forecasting method — weighted pipeline by stage — breaks down and has to be replaced with something else.

3 — Forecasting

Unlike SaaS forecasting, which is a judgment exercise on which deals will close and which will slip, consumption forecasting is closer to an analytical modeling exercise. Most of your revenue for the quarter is already in motion. The question is whether you can accurately read the trajectory.

In a consumption model, most of your revenue for the quarter is already locked in — it comes from existing customers who are already consuming. The forecast question is really: what will existing customers consume this quarter, and how much incremental consumption will sales drive on top of that?

At MongoDB, we built a model that tracked weekly growth trends, updated every week throughout the quarter. We then built a sensitivity analysis for the remaining weeks showing where we might land depending on whether growth held, accelerated, or slowed. Here is an illustrative example:

We then layered in insights from the field. Sales would flag any outliers the data model couldn’t pick up, such as a customer doing a major spend optimization or bringing a massive new application online.

One challenge we didn’t fully solve: cascading this methodology down to the field. At the aggregate level, the model was fairly accurate. But once you zoomed into specific regions or smaller books of business with fewer customers, the statistical smoothing broke down.

Wrapping Up

Pipeline and forecasting present major challenges for companies that have usage-based pricing models. But in some ways, the harder problem comes next: how do you give a rep a number to hit when their customers’ consumption can swing up or down based on factors outside anyone’s control? That’s what I am working on for my next post.

In particular, she was tasked with improving the workflow for the Customer Success team. They were using multiple SaaS tools, including Gainsight, Salesforce, Tableau, and Terret. There were also custom, internal tools purpose-built for the CSMs.

Yet despite all of the insight these tools were supposed to provide, the CSMs were frustrated. They were opening five or more tabs to answer a single customer question. The “call to action” system fired off so many alerts and created so much alert fatigue that people had basically stopped looking at it. The result was tool sprawl that made simple tasks harder, not easier.

How tool sprawl happens

Tool sprawl is a common pattern at fast-growing organizations, and it’s a pattern that new AI vendors and vibe-coded apps will only accelerate.

The first step in Nina’s process to untangle these systems was interviewing CSMs about their day-to-day workflows. No single tool had the full picture. A CSM trying to understand a customer had to stitch it together across systems: checking health scores in Gainsight, pulling account history from Salesforce, looking at usage trends in Tableau, and cross-referencing signals in the internal tools.

How did this happen? Different stakeholders were responsible for different metrics. For example, the product management team wanted to understand feature adoption. The finance and sales operations team wanted to understand the forecast. The CSM leadership wanted to understand onboarding and measure customer health.

The result was that each of those teams had introduced tools in isolation, creating a Frankenstein of systems. Nobody had deliberately designed this arrangement. It had evolved naturally as the company grew and different teams made different tooling decisions to track their own metrics. (And for the record, I am 100% sure that I personally contributed to this tool sprawl!)

“The systems reflected our internal org chart, instead of the end users’ needs,” Nina explained.

It’s not that anyone made bad decisions. Every team made reasonable decisions for their own needs. The problem is that “reasonable for my team” and “coherent for the end user” are two different things.

Buy-in has to go bottom-up

To kick things off, Nina’s team ran a cross-functional workshop before making any major decisions. Day one was tool walkthroughs and a SWOT analysis. Day two was alignment on the actual problem statement. She brought together people from different orgs who had never been in the same room talking about the same problems.

This exercise was eye-opening for all the different groups that had rolled out tools to the CSMs. They could now see and understand the tool fatigue the CSMs were experiencing, and were motivated to address the end-to-end process instead of their individual needs.

When it came time to start designing the systems, Nina’s took a bottom-up approach to getting buy-in. The end users had the biggest pains as a result of the tool sprawl. Redesigning the process from their perspective helped Nina identify and recommend parts of the workflows that could be removed or simplified. By the time she met with the executives, she was reporting a decision that’s already been validated.

The goal was a coherent workflow (not just “fewer tools”)

It is tempting to start with the solution: let’s consolidate systems. Some of Nina’s stakeholders suggested a “lift and shift” of functionality from Gainsight into Salesforce to reduce the number of screens needed. A lift-and-shift would successfully reduce the number of tools, but it would have replicated many of the old problems in a new environment.

Instead, Nina mapped what the onboarding process was supposed to accomplish. She identified which of the 15 steps in the existing checklist were necessary versus administrative overhead that had accumulated over time. And with lots of back and forth between the CSMs and their management, she was able to cut that process to five steps.

She knew the team wanted fewer tools, so she investigated whether the new onboarding workflow could be implemented in Salesforce. Her team ultimately de-commissions Gainsight and rolled out the streamlined process directly in Salesforce. But she didn’t enter the project with a goal of eliminating a particular tool. She entered the project trying to understand the process and which metrics were critical to track.

What happens when you get the process right first

The onboarding improvements were one of many major changes Nina’s team implemented.

She also rebuilt the call-to-action system that was firing too many false positives, not by buying a new tool, but by fixing the underlying data science model triggering those calls-to-action. She was able to reduce from 30 CTA types to 8.

She redesigned the forecasting process so that using Terret was the natural thing to do, not the extra thing. Terret adoption went from 20% to over 80%.

Nina piloted each of these changes before full deployment, reporting back to the CSM management on feedback and issues at each step.

The lesson: it’s the fool, not the tool

With so many new AI tools on the market, it’s tempting to keep bolting on new systems. But AI doesn’t fix broken workflows. It just helps you execute them faster. Teams that skip the hard work of talking to end users and mapping the end-to-end process will find themselves in the same place they started.

Hey, these challenges will keep us ops people employed… at least, for now 🙃

I spent eight years running Sales Operations at MongoDB, including the years when we made the transition from subscription software to a consumption model. I wrote this article for the founders and GTM leaders navigating this now, particularly those building AI products where per-token or per-API-call pricing is the default. If you charge on usage, you need to think about comp on usage. Here’s everything I learned.

TL;DR:

• If you’re building an AI business, it’s likely priced based on usage. The pay-per-use model pioneered by AWS is now standard for AI-native companies: per-token, per-API-call, or hybrid seat-plus-usage.

• Traditional SaaS comp breaks in a consumption world. When revenue is usage-driven, the deal close is just the starting line. Measuring rep impact, setting fair quotas, and avoiding perverse incentives all require fresh thinking.

• There are three main models: paying on committed spend (similar to traditional SaaS), paying on incremental usage, paying on gross usage, or some combination of those three. Each has trade-offs.

• I share my personal experience iterating on the MongoDB compensation plans when I was head of Sales Operations. We used the incremental usage model, and fine-tuned it over the years.

What do we mean by a “consumption” or “usage” based pricing model?

Before we get into the compensation mechanics, it’s worth grounding ourselves in what makes a consumption model fundamentally different from what came before it. Let’s explore the three eras of software pricing:

The perpetual license era (think Oracle) meant the customer bought the software outright, typically via large upfront deals. The vendor owned the market, pricing was complex, and the customer was locked in. This was a very vendor-friendly world.

The SaaS era (think Salesforce) shifted toward subscription-based pricing where the customer “rents” the software and renews annually. More customer-friendly, but still with friction, including large upfront commitments and complex pricing tiers. And from a metrics standpoint, you have a misalignment between bookings and actual revenue recognition. (Bookings are recognized upfront while the revenue is recognized ratably over the course of the contract.)

The consumption era (pioneered by AWS and now AI-native startups) means that the customer pays for what they use, billed in arrears. Usually paired with a product-led growth and/or self serve model, the product has to continue to prove itself in every single billing cycle.

“For customers, the benefits of doing business with vendors that offer a usage-based pricing model are clear: use the products you want, when you want, and be billed when you do.”

— Ben Chambers, Sales Compensation Consultant (formerly New Relic, Databricks)

This pricing model is particularly relevant in the age of AI. Many AI native companies price on a per API call or per token basis, or use a hybrid of per seat models with usage. If you’re building a product that calls an LLM under the hood, your pricing model is almost certainly consumption-based in some form, which means your sales comp model needs to be too.

During my tenure leading Sales Operations at MongoDB, we shifted our pricing and sales motion from subscription to consumption based. In this article, I’ll share the evolution of that model over time and compare it to other companies that I studied while designing the plans, such as AWS, Twilio, Snowflake, and others. These insights should be applicable for founders and GTM leaders at AI companies who are navigating building their business around usage-based pricing.

The two main usage-based compensation models: incremental usage and gross usage

In working on MongoDB’s shift to consumption, I researched several different approaches and spoke with RevOps leaders at many companies. The compensation models typically fell into the following categories:

1. Pay on committed spend, similar to a traditional SaaS model

2. Pay on incremental usage (example: MongoDB)

3. Pay on gross usage (example: AWS)

4. Some combination of the above, e.g. a comp plan with an element for committed spend as well as usage

I’ll start with the incremental approach, since that is what we adopted at MongoDB. We didn’t get there overnight, so let me walk you through the process.

Compensating on incremental usage

MongoDB built their sales incentives around a metric we called Net ARR. I want to share how we actually got there at MongoDB, because it took several iterations on the plans over many years, each one fixing a problem the previous stage created.

Stage 1: Paying on commitment

When MongoDB first introduced its cloud product, Atlas, in 2016, the company was operating in a world built around traditional software sales. Sales had been selling an on-prem product (MongoDB Enterprise Advanced) where they were credited using a traditional New Annual Contract Value (NACV) method. Atlas, in contrast, would be billed based on usage.

When we introduced Atlas, we initially applied the same NACV logic to sales compensation. Reps were credited when they got the customer to agree to purchase a predefined pool of credits, from which the customer would draw down as they used the product. When they ran out of credits, they would have to go back to sales to either renew early or do a short “top-up” deal to bridge their usage until the end of the contract.

Challenges with this approach:

• It introduced friction into the sales cycle. Customers had been conditioned by the hyperscalers to pay on a usage basis and had come to expect that pricing model.

• To purchase a commit, the sales would need to go through a sizing exercise to figure out how much the customer should buy. Sales engineers were spending weeks trying to help the customer predict their spend.

• There was an incentive for reps to pressure customers into commits before they were ready. Even worse, in some cases they might oversell the product, getting customers to make large commitments and then walk away before the customer implemented the technology.

• Reps were wasting cycles chasing paperwork for “top-up” deals with customers once they ran out of credits. This was annoying for the customer too, most of which preferred to be billed overages in arrears rather than sign a new contract.

Stage 2: Neutralizing the incentives with run rate crediting

The goal of Stage 2 was neutralization: we wanted reps to be indifferent between selling an Enterprise Advanced commit, an Atlas commit, or having a customer pay-as-they-go. Whatever was right for the customer should be right for the rep.

After a customer consumed their commit, we would bill in arrears and pay reps on the incremental run rate above the commit. This solved the problem of reps chasing top-up deals.

Once the customer burned through their commitment, the rep would automatically get credited for any incremental ARR. And if the customer had no commitment at all, we’d pay on the run rate directly. This addressed the issue of reps prematurely pushing commitments, and also reduced (but didn’t eliminate) the time spent working on sizing deals.

Example

Let me illustrate it for you in an example.

• Q1: The customer commits to 120 and they use 25. That means that their annualized run rate is 100 (25 per quarter x 4 quarters). The rep would get credited for the commit of 120.

• Q2: The customer starts to ramp up their usage and consumes 35. This means that their annualized run rate is 140, exceeding their original commit. However, because they have only consumed a total of 60 (25 in Q1 and 35 in Q2), the rep wouldn’t earn any incremental credit.

• Q3: The customer’s usage accelerates further and they consume 50. This brings their ARR to 200 (50 per quarter x 4 quarters), and they have used up their commit of 120 (25 in Q1, plus 35 in Q2, plus 50 in Q3 is 135). The customer wouldn’t need to go through procurement again; they would start getting billed for the incremental usage. And the rep would get credited 80, which represents the difference between their baseline of 120 (established by the original commit) and their run rate (200).

• Q4: The baseline on the account would be 200. In order for the rep to earn further credit, the customer would need to increase their usage or commit to more than 200.

What worked:

• Reps were still incentivized to drive commitments, but they weren’t pushing them on customers unnaturally.

• Sales and the customer spent less time on sizing commitments.

• Reps (and customers) no longer had to do top-up deals.

Challenges with this approach:

• We started to notice reps racing to pull renewal commitments forward at the end of quarters, and sometimes even sacrificing discounts, in cases where a customer’s run rate exceeded their commit but they hadn’t yet exhausted the commit balance.

• When a customer’s use declined below the baseline, reps were subject to commissions and crediting clawbacks.

Stage 3: Greater of the run rate or the commit

To address the issue of reps pulling forward renewals, we made a simple change to the crediting guidelines. The rep would be paid on the greater of the customer’s ARR run rate or their commit, whichever is higher, regardless of whether they had exhausted their commitment.

If we revisit the example above, under the new crediting rules, the rep would have earned credit of 20 in Q2 because their run rate (140) exceeded the baseline set by the commit (120) even though the customer hadn’t exhausted their commit. This effectively pulled forward ARR crediting, and had to be taken into consideration in quota setting. (Said another way: quotas went up because reps could get ARR more quickly and easily.)

What worked:

• Reps no longer pushed customers to renew early if they hadn’t exhausted their commitment, putting them in a better negotiating position.

Challenges with this approach:

• We continued to have challenges with customer churn impacting reps’ paychecks. This was largely addressed through an exception process that was highly manual, and was the major pain point of this comp plan.

Compensating on gross usage

While MongoDB was undergoing this transformation, I spoke with peers at many other companies, and most had some element of compensation on gross usage. The logic is intuitive: every dollar the customer consumes is a dollar of revenue for the company, so give the rep credit for every dollar billed. The quota is set as a total billing target, factoring in an expected organic growth rate, plus an additional uplift that the reps were responsible for driving.

Here’s a simple example of how you might set a quota in this model:

• Territory billed $1,00,000 in the prior year

• Organic growth is 25% → $1,250,000 of billing expected

• Quota is set with an uplift of 20% → $1,500,000 quota

What works:

• The elegant part of this model is the alignment: the customer, the company, and the rep are all measuring the same thing. Unlike the incremental crediting approach, the crediting model is simple, and you don’t have to clawback if a customer’s usage drops. (The reps just earn less of their overall variable and have to make it up elsewhere.)

• Reps have predictable payouts as customers consume.

• This can be paired with a separate element for commitments. In some cases, prior year commits informed the consumption element of the comp plan, and the commit element would be weighted (e.g. 30% on commitment, 70% on usage) based on how many customers were available for renewal.

Challenges with this approach:

• The biggest challenge is quota-setting. You need a model to project growth for all of your accounts, and these growth expectations can feel like a black box to reps.

• Reps earn commissions on the usage across their entire territory, including the portion that would have grown without them. You’re effectively paying on retention as well as expansion. And you can suffer from the “free rider” effect, where reps can earn a large portion of their variable just for retaining the customer.

• It’s harder to make an attractive comp plan with big upside, because the quotas end up being so large.

There’s no perfect comp plan for consumption

When I’m asked about compensation plans for consumption businesses, my answer is always the same: There is no perfect solution, only trade-offs. I’ve described many above, but here is the cheat sheet:

If you’re building an AI company on usage-based pricing, expect to start somewhere imperfect and improve from there. The goal isn’t to design the ideal plan on day one. It’s to design a plan that you can measure, that reps can understand, and that you can change when the incentives drift from the behavior you actually want.

Sources & Related Reads

• The AI pricing and monetization playbook from Bessemer Venture Partners

• A new framework for AI agent pricing from Growth Unhinged

• How Do I Build a Usage-based Sales Compensation Plan? from the A16Z blog, co-authored by my former colleague Joe Morrissey

• Overages: The Overlooked Problem in Usage-based Pricing (and How to Fix It) from the A16Z blog, also by Joe Morrissey

• Navigating Consumption-Based Pricing with Mike Scarpelli, CFO of Snowflake

• Sales Compensation in a Consumption Pricing World, Andrew Straus on the Snowflake blog

• Going Usage-Based? It’s Time to Rethink your Sales Comp by Ben Chambers

• The Cloud Consumption-Based Model: Changes Sales Leaders Can Make to Ensure Success, Laura Palmer on LinkedIn

• Usage-Based Pricing: Behind the Scenes of New Relic’s Transformation, OpenView blog

• Google Cloud Considers Scrapping Sales Commissions Based on Deal Values, The Information

• The 10 Commandments of Sales Compensation

• Finally, embedded below is a talk I gave at SaaStr on this topic in 2023

The failure usually shows up in one of two ways:

1. The founder hires a marketer to focus on messaging, and then disappears, assuming positioning is now “handled.” It isn’t. Messaging requires the founder’s ongoing involvement, especially when the product is still evolving.

2. The founder hires a marketer to drive pipeline before they have product-market fit or positioning. Spending $$ on distributing a sub-optimal message is $$ down the drain.

Marketing is different from almost every other function you’ll hire for. It’s a function with several distinct sub-disciplines.

In sales, the rep in New York is doing the same job as the rep in London. In engineering, yes, you have specialties, but a senior engineer in one stack can usually ramp on another. Marketing doesn’t work that way.

Under the marketing umbrella you have product marketing, communications, demand generation, marketing ops, and design. These are distinct skill sets. Hiring a great demand gen person and expecting them to also own your messaging strategy is like hiring a great radiologist and expecting them to see patients in the emergency room. They’re both doctors. One diagnoses; the other treats. You would not want them to switch.

So when a founder asks me “should I hire a VP or a more junior do-er?” I usually push back on the question. The more important question is: which flavor of marketing do you actually need right now?

AI has changed this a bit. Senior marketers can do more (and should be expected to do more) than ever before. One person with the right instincts and good AI tooling can cover ground that used to require a team. But it hasn’t collapsed all the disciplines into one. It’s mostly made good marketers faster.

Product marketing vs Demand generation

For most early-stage B2B startups, it usually comes down to two options: someone with a product marketing background, or someone with a demand generation background.

The case for starting with product marketing

Founders are often better at positioning than they realize. You sold the early deals. You know the domain. You’ve had the customer conversations. The problem isn’t that there’s no story, it’s that the story lives in your head and nowhere else.

Take Charity Majors, the founder and CTO of Honeycomb, as an example. Charity didn’t just build a product. She invented a category, using the term observability as early as 2016 to describe and differentiate Honeycomb from simplistic monitoring tools.

That’s product marketing at its best, and it happened because someone with deep domain knowledge spent lots of time thinking about the problem, speaking with customers, and even writing an entire book on the topic.

You may think that I’m making the case for founders to lead product marketing. And I do think founders are best positioned to drive messaging. But hiring a product marketer can make sense as long as the founder stays involved. For example:

• A product marketer can help get the story out of the founder’s head and into a format the rest of the company can actually use: website copy, sales decks, one-pagers, competitive battlecards.

• A product marketing hire is a good fit if your product is evolving quickly. In the early days, when what you’re selling can change substantially from quarter to quarter, a product marketer can own the process of keeping messaging up to date and aligned across all the teams that are customer-facing.

• A product marketer is also a great fit if you are scaling your sales team. Sales will demand sales collateral and provide constant feedback on positioning based on their customer conversations.

The biggest failure mode is the founder who hands off positioning and then disappears. Messaging is not a one-time exercise. The market evolves, competitors shift, you learn things from customer conversations that should feed back into the story.

Charity was inspired by the term observability as it was being used in another domain (hardware) and then “spent a couple of years trying to work out how that definition might apply to software systems.” Yes, it took years for her to nail the messaging. So don’t expect to solve positioning over night.

The case for starting with demand generation (or a growth marketer)

If product marketing is about what you say, demand generation is about getting people to actually hear it. Running campaigns, testing channels, driving traffic to your website, and filling the funnel is the job of a demand gen marketer. Content, events, webinars, digital channels, whatever combination works.

Some founders take this role a step further and hire a growth marketer. In addition to running campaigns, a growth marketer will also be working on optimizing the product itself, seeing the first-time user experience as an extension of the marketing funnel. This profile typically makes sense in a company with a product-led growth approach.

The best growth marketers share a trait: they experiment constantly, kill what doesn’t work fast, and double down hard when something does. They are allergic to running the same playbook quarter after quarter regardless of results.

Product marketing builds the message, and demand generation distributes it. For that reason, demand generation typically builds on product marketing.

This is the right hire if:

• Your messaging is somewhat stable you’ve got product-market fit, customers who get what you do and love it, but you need more of them to know you exist, and/or

• You’re a founder who stays deeply hands-on with messaging and needs someone to build and run the distribution engine.

So which profile should I hire?

Start with the problem. What’s actually blocking you? If everyone’s confused about what you do, hire a product marketer to refine the message. If everyone who hears about you loves it but almost nobody has heard of you, hire a growth marketer to drive awareness.

And if someone suggests you hire a VP who will “oversee all of marketing” before you even have a team, just know that what you’re actually getting is someone who manages. What you need, at this stage, is someone who does things.

More on this topic

• I recently co-authored a post for the Ansa blog When (and how) should I make my first marketing hire? with my long time friend and former MongoDB colleague Justin Dunham. Justin now runs an agency called ércule that helps startups build organic and content marketing. Watch the Ansa blog as this is the first in a series on this topic.

• Ansa and Base10 are co-hosting a breakfast event for founders at Base10’s San Francisco offices, where Justin and I will be talking about this topic, along with Austin Lau from Anthropic. You can apply to attend here.

“We need to slow down hiring,” the ops team told me.

“Are you sure that’s the right conclusion?” I asked.

“We need to gain more confidence that our AEs can attain their targets before investing more,” they explained.

“But do you know if you missed your number due to a productivity issue or a productive capacity issue?” I challenged.

“Well… it’s both… I think.”

Productive capacity and productivity are the most fundamental metrics in a sales organization. Not only do they help you forecast your expected output, but examining them is the first step to understanding how your sales organization is performing.

Yet, these two terms are frequently confused and conflated. So let’s get precise about what these two things actually mean.

Productive Capacity vs Productivity

Productive capacity is a projection of what you think a team should produce in a given period. It’s a representation of how much “firepower” you have. If a ramped rep produces, on average, $1M per year, and you have 5 fully ramped reps, then you have $5M in productive capacity (5 reps x $1M/rep per year).

Productivity is your average output per rep. It’s an efficiency metric that normalizes performance on a per capita basis. If you have 5 reps and they collectively produce $4M in a year, your productivity is $800K per rep ($4M / 5 reps = $800K per rep).

Here’s a quick example:

• Team A: 12 ramped reps close $12.6M in Q2 → $1.05M productivity per rep

• Team B: 30 ramped reps close $21M in Q2 → $700K productivity per rep

On the surface, Team B looks like the better team. They closed $8.4M more. But on a per-rep basis, Team A is dramatically outperforming.

A Quick but Important Detour: Quota Is Not Productivity

Before we go further, it’s worth pausing to discuss a pet peeve of mine: when people use productivity and quota interchangeably.

Quota is how you pay people. Productivity is what you actually expect them to produce. They’re related, but they’re not the same thing.

Most companies set quotas above expected productivity deliberately, by 15-25%, to create a performance buffer. So when you’re modeling what a team will actually deliver, you shouldn’t be using quota as your input, you should be using historic productivity. And ideally, you are using different productivity assumptions for each segment and geography.

Here’s why it matters: if you have 10 sellers each carrying a $2M quota, you have $20M in “street quota.” That sounds like a lot of firepower. But if those reps realistically produce $1M each based on historical performance, your true productive capacity is $10M. If you only look at the quota in the field, you’ve just inflated your sense of safety by 2x.

Here’s another reason it matters. When a new rep joins, they typically have a quota from day 1. It might be a lower quota to reflect their ramp, but they have a quota nonetheless. Yet, it’s unrealistic to expect a rep to start producing immediately. For this reason, I would typically model productive capacity at $0 during the ramp period.

Why This Distinction Matters

Going back to the company I was advising: it’s entirely possible to miss a revenue target because of rep attrition, while the remaining reps are overachieving. In that scenario, the right answer might be to double down on hiring, not pull back. (And to diagnose why you are having such attrition.) If you don’t separate the capacity question from the productivity question, you won’t see it.

The first step in diagnosing a performance problem is to break it into one of these two categories:

• Capacity problem: You don’t have enough reps. The team is producing well per capita; you just don’t have enough of them.

• Productivity problem: You have plenty of reps but on average they are underperforming. Adding more headcount here just scales the problem.

An important callout: as your sales organization gets larger, it becomes as important to forecast hiring and attrition as it is to forecast revenue. But that is a topic for another article.

If it’s a capacity problem, then you can narrow in further on the issue:

• You aren’t hiring fast enough

• You are attriting too many reps

If it’s a productivity problem, it will take you on a different path, where you will want to look at:

• Productivity by segment, geography, and tenure cohort

• Pipeline generation (is it a top of funnel issue?)

• Close rates (is it a deal execution issue?)

• Ramp times (are you hiring the wrong profile or not enabling/onboarding?)

This flow chart illustrates the process:

If you enjoyed this topic, I recommend reading The Use of Ramped Rep Equivalents (RREs) in Sales Analytics and Modeling by my favorite blogger Dave Kellogg.

I laughed, because Oso’s marketing budget at the time was a rounding error on an enterprise marketing budget. And yet here was a founder of a nearly-billion-dollar company getting stalked across the internet by their ads. (And yes, this company was in Oso’s ICP!)

The combination of AI-powered tools and data has enabled small teams to execute sophisticated campaigns on a shoestring. This is a case study on how Ege Ayan, Oso’s digital marketer, ran that very campaign.

“When you’re small, you need to be precise,” Ege explained to me, “A broad campaign can burn through a limited budget.”

Instead, when you’re small, you have to be surgical, by using data to find exactly the right people and follow them around like you have a much bigger budget than you actually do.

Good creative doesn’t need to cost a fortune

Before you can target anyone, you need something worth targeting them with. Ege wanted a polished video ad that would appeal to developers while also signaling the maturity of the company and product. Without a dedicated creative team or budget for an agency, Ege had to be scrappy.

Ege sat with Oso’s DevRel team to brainstorm on several concepts. After many iterations, he came up with the concept of showing a developer adding a role — the kind of thing that’s genuinely painful without a good authorization solution — and landed on a punchy call to action: Just call Oso.

Once he had nailed the messaging, he created a detailed creative brief with frame-by-frame instructions. By the time he found an animator on Upwork, that animator could focus on execution rather than strategy.

Total cost: under $1,000.

In this case, Ege outsourced the execution in a cost-effective way. But with AI video tools like HeyGen, it’s also possible for marketers to produce professional video content without outsourcing at all.

Get precise about targeting

The video was the bait. But bait only works if you put it in the right water.

Now that Ege had an ad that addressed a specific pain point, he set out to find the audience to target. Here is how he got the ad to follow that founder I was talking to all over the internet.

Step 1: Find the companies with the pain using Sumble

Ege started with Sumble. Sumble analyzes company websites, job listings, and people profiles to figure out their tech stack and key technology initiatives. Using Sumble, you can use natural language queries like “find me companies with over 1000 employees who are using Apache Kafka in the east coast of the US” to build lists of accounts to target.

Ege used Sumble to identify companies actively recruiting for roles that had terms like roles, role based access control, authorization, or something similar. These would be prime targets for this campaign.

Step 2: Enrich the company list with Clay

Ege then fed those Sumble accounts into Clay and ran them through Claygent (Clay’s AI agent) with custom prompts to further enrich and qualify each account.

“They were basically separate angles to determine whether they have complex authorization,” Ege explained. “It would check for evidence of multiple roles in their app, custom roles, or fine-grained access control.”

Step 3: Find the right people on all social media platforms, including Meta

While Ege could build audiences on LinkedIn targeting specific companies, he wanted to see if he could get better ROI by advertising on Facebook and Instagram. He hypothesized that he would get a better response on those personal platforms than on LinkedIn, which is flooded with corporate ads.

The challenge? Most data enrichment provides corporate email addresses. To match people on Meta, he would need personal data.

(Side note: I have to laugh at the irony of looking for personal emails for B2B marketing, as I had spent years at MongoDB trying to figure out the opposite problem: what the hell to do with the thousands of gmail, yahoo, and other personal emails we collected through Atlas cloud sign ups.)

Ege started by using Clay to identify personal emails. This quickly got expensive and he was burning through Oso’s limited pool of Clay credits. So he pivoted to using an enrichment provider called Primer, which specializes in helping marketers find B2B audiences on platforms like Facebook and Instagram.

Step 4: Experiment with different channels, even controversial ones

Ege wanted to take his “target a B2B audience on B2C channels” approach a step further, and he decided to experiment with Google Performance Max.

If you ask ChatGPT what marketers think of Google PMax, you’ll probably get answer along the lines of: Marketers have a very mixed (often polarized) view of Google Performance Max (PMax). It’s one of those tools where people either love the results—or deeply distrust how it gets them.

But Ege’s view was that the usual objections assumed you were going in blind. He wasn’t. With a clean, Primer-built audience already in hand, he had something most PMax campaigns lack: a precise list of who he actually wanted to reach. That de-risked the channel’s biggest weakness. He set aside some experimental budget to see if he could get results from this channel.

The results

Across the board, the campaigns delivered lower CPCs and higher click-through rates than typical broad targeting, a direct reflection of how much more relevant the ads were to the people seeing them. The Sumble-sourced accounts, in particular, outperformed standard retargeting, which makes sense. Someone actively hiring to solve an authorization problem is a warmer prospect than someone who once clicked on your website.

PMax was messier, as expected. Conversion rates were lower and yes, there was some junk traffic. But Ege did the math: the volume was high enough, and enough of those conversions were real opportunities, that the channel earned its place in the mix.

And then there is the benefit that you can’t measure. Let’s go back to the story at the start of this post. That founder, running a company approaching $1 billion in revenue, with a full marketing team, thought Oso was everywhere. He felt outspent. He wasn’t. He was just out-targeted.

That’s the whole game. You don’t need to be everywhere. You need to be everywhere the right people are looking. With the right data and the right tools, a one-person marketing team can pull that off.

Appendix: The Tools

For anyone who wants to replicate this playbook, here’s what Ege used and what each tool does:

• Upwork & HeyGen: Ege found the animator for the video on Upwork, which is a good reminder that great creative doesn’t require an agency. You could also explore tools like HeyGen to make AI-generated video content. (I’ve written about using HeyGen in a previous post.)

• Sumble: Surfaces company hiring data as an intent signal. If a company is recruiting for roles that indicate a specific pain point, that’s a warm signal worth targeting.

• Clay: Data enrichment and list building. Pull in a list of target accounts and enrich them with firmographic and contact data at scale. Feed Claygent a prompt and a list of accounts, and it will research and qualify them automatically.

• Primer: Takes your account list and finds the right contacts within those accounts to match against ad platforms. Particularly useful for building LinkedIn and Facebook custom audiences without relying on personal email matching.

1. Know what you are optimizing for.

This may seem obvious, but it’s worth stating explicitly. The comp plan communicates the behaviors that you want.

It’s easy to skip this step because, doesn’t everyone want more revenue? Well, not all revenue is created equal. Do you value new logos over retention and expansion of existing customers? Are there certain market segments that are a better fit for your product? If you have a product-led growth approach with an option for customers to purchase self-serve, do you want reps calling accounts that have started spending on a credit card, or is it more efficient to leave them in the self serve channel?

You need to start with these questions, before putting anything into a spreadsheet.

2. Not everything can be managed in the comp plan.

Compensation is a powerful lever, but it’s not the only one. Inspection, management, and coaching all matter too.

Take pipeline generation: an AE needs to build pipeline to hit an ARR target, but that doesn’t mean you need to put pipeline metrics into the comp plan. Build dashboards. Inspect. Coach.

Another relevant example: Sales Development Reps (SDRs) are the most junior reps in the organization responsible for building pipeline. They are highly motivated by career trajectory, in the hopes that they will graduate into a quota carrying role. They will pay close attention to criteria for promotions, often more than they focus on their comp plan.

Trying to shove everything into the comp plan is the best way to over-complicate your plan, and confuse the hell out of your reps.

3. Upside and risk go hand-in-hand.

Sales engineers, channel reps, and SDRs sometimes push back on how much more AEs earn. Why do they earn so much, when they didn’t find the deal or didn’t do the technical demo?

Because the risk profile is completely different. Typically 50% of an AE’s comp is variable, and their job is on the line every quarter. The supporting roles carry less upside, but also far less risk. That’s not an accident.

What does this mean practically? For AEs, you want to give them a path to at least double their variable (and hopefully more). In contrast, the supporting team members should have much more modest upside. If they want the big bucks, they have to go carry the bag.

4. Beware of blunt instruments.

When you need to contain costs, it can be tempting to cap commissions. Resist that urge. If a rep hits their cap, you’re telling them to go on vacation for the rest of the year. It’s also a strong signal to your best performers that there’s a ceiling on how much they can earn.

There are other ways to manage cost structure without punishing success. You can decelerate past a certain point, i.e. at 300% of plan, the reps rate goes back to their base rate instead of the accelerated one. You can incorporate a gate to accelerators that makes sure that the overachievement happens the right way (e.g. you must sell professional services with the software to make sure that the customer implements). You can add terms in the comp plan agreement to do the same, e.g. you can’t earn more than your OTE on a deal that you inherited from another rep. And I’m sure there are others that I’m not thinking of.

5. Don’t be penny-wise and pound foolish.

Speaking of upside: offer aggressive accelerators for your AEs. Especially in the early days, outsized performance can have a meaningful impact on overall revenue. At the early stage, when driving growth is existential, why not reward handsomely?

I recently recommended top-tier accelerators of over 25%. The founder responded, “but is it okay to pay that much on a deal?” My answer: if they bring in $5 million of incremental revenue, that’s a very happy problem.

Sales tends to be outlier driven. A small percentage of reps can and should get those big paychecks, and those are the ones you definitely want to retain.

6. Keep it simple.

Compensation drives behavior, but only if people understand what you’re asking of them. The person receiving the plan should be able to look at it and immediately know what they need to do.

Everyone aims to keep comp plans simple, yet rarely succeeds. Here are a few ways they can get complicated quickly:

• If you have four components to your comp plan, it won’t be clear what to focus on. And the rep’s variable will be so diluted that it spreads the rep thin. (I’d recommend no more than two or maybe three elements, an “element” being something you’d quote and pay on, like ARR or services.)

• Splitting or reducing credit (e.g. if the deal is sourced by the channel the rep gets x% of the deal…) creates both confusion and friction between teams.

• Too many bells and whistles, such as multiple gates or multiple accelerated elements.

• Terms and conditions that require a law degree to understand.

7. The devil is in the details.

But, Meghan, you just told me to keep it simple? And now you are telling me to focus on the details?

In a comp plan, the details really matter. How do you pro-rate quota for someone who changes roles mid-year, or joins in October? Do you increase quotas across all currencies by 5%, or set USD as the base and convert FX? Which exchange rate do you use, and when? Do commissions pay at booking or at cash collection?

These micro-decisions add up fast, and getting them wrong is expensive. Discuss, debate, and document everything. And then you need to pick your battles, and decide which items you want to explicitly cover in the comp plan. You may even leave some items sub-optimal for the sake of simplicity.

8. Quota-setting is an art and a science.

The science: model quota coverage top-down and track your “street quota.” Top-down modeling simply means building a capacity model with the number of reps you plan to hire and your expected attrition rate. Your street quota is the every-changing sum of the quota you’ve actually assigned. It will incorporate the actual hiring and actual attrition, as well as any exceptions that you make.

The art: overall coverage is a judgment call, not a formula. Typically you want more quota on the street than your company target, but how much more quota? No spreadsheet will tell you the right answer. You need to balance making sure you have enough quota on the street to hit the plan, without making the target so unrealistic that your reps all quit (which, of course, reduces your street quota). And you need to constantly track the street quota against your plan.

9. This is a test to see if you are reading this article.

I had a bunch of candidates for the 9th commandment, including:

• Remember that people hate to lose more than they love to win

• Start by defining what success looks like, as well as what excellence looks like

• Reps are like water, they flow to the path of least resistance

But I couldn’t really settle on one, so I thought I’d add this Easter egg instead. Make a note in the comments to share which principle you think I am missing.

10. Sell the plan.

Designing a great comp plan is only half the job. You have to tell a story: not just here’s the plan, but here’s why it’s designed this way and here’s how you can make a lot of money.

Before rolling it out broadly, I always previewed it with a few people to gauge their reactions and get ahead of objections. And I’d ask the most aggressive reps how they’d “break” it. If they can find the loophole, your reps will too, and it’s better to find out first and prepare.

You need to enable the reps and their managers on the benefits of the plan. A good plan can be a recruiting tool, and the managers need to know it inside and out. And when you roll out the plan at the beginning of the year, show them the scenarios: here’s what you need to do to earn twice your OTE. Here is your path (even if highly unlikely) to earning a million dollars.

The plan is the strategy. How you sell it is the execution.

Acknowledgements: Thanks to KJ Logue for his expertise on comp planning and Graham Neray for edits and feedback.

While the latter half of my 15 year stint at MongoDB was running sales operations, I got my start as a generalist, largely focused on marketing and developer relations. You could say I was the “first GTM hire” but people didn’t really use the term “GTM” for sales and marketing back in 2010. And it was such a small team (less than 10 people when I joined) that we all kind of did a bit of everything. More on that later when we talk about conferences and events.

While it’s kind of Tim to say that I had a hand in blazing the DevRel trail, I often think that I get too much credit for the early success. It really started with the product.

The OG PLG

Dwight and Eliot made an early decision that became the basis for the company’s success: they made MongoDB open source.

Another term that didn’t exist back in 2010 was product-led growth (PLG), the idea of building your business with a product that delights customers and can be accessed without the clunky interface of a sales rep. In the years before MongoDB launched a cloud product, open source was the equivalent of a freemium model.

Downloading the software was free and completely frictionless – we didn’t even require any information from developers to download it. (The fact that downloads weren’t gated behind an email form would be cause for much consternation among the sales team, who complained that they didn’t have visibility into the people using the software. A problem that I was later faced with while running sales operations.)

At the heart of the viral adoption of MongoDB was the simplicity of the product. When I spoke to developers during many long days of booth duty at events like PyCon or JavaWorld, they spoke enthusiastically about how intuitive MongoDB was. Organizing data in tables and rows was complex, and didn’t map to the way that they wrote code. Storing data in JSON documents felt natural.

(Few people know it, but the leaf in the MongoDB logo embodied that intuitive user experience. Dwight gave the designer instructions that he wanted something that felt “natural.”)

The first-time experience was critical. Not only did it need to be easy to get up and running, but if you needed help, Eliot would typically respond to questions on the user forum in less than 15 minutes. “We can’t give people a good experience,” I remember Dwight saying, “it needs to be amazing, so that they remember and tell their friends about it.”

This product experience is the foundation for the early success of MongoDB. Making developers feel like they were part of a movement, not just users of a tool, simply added fuel to the fire.

The in-person developer community

I took direction from Dwight. Even though he was CEO, he spent most of his time coding, and he understood developers.

I was struck by Dwight’s comment about giving people an amazing experience. When I saw community members answering questions on the forums, posting on social media, or writing about MongoDB, I would reach out to thank them. Soon I was mailing MongoDB coffee mugs all over the world.

We made a list of developer events, conferences, and meetups, and started reaching out to organizers to get a MongoDB engineer on the agenda. Soon the small team of engineers (remember we were less than 10 people at this point) were getting pulled into events so much that it was eating away at their programming time. I quickly started tapping that list of MongoDB enthusiasts I had mailed mugs to, offering to sponsor them to attend a conference and speak about MongoDB. Honestly, what now seems like a brilliant hack was born out of necessity: the engineers were frustrated that events were eating into their coding time, and I was caught between Dwight’s desire to be everywhere and a team that just wanted to code. But this approach ultimately made the community an extension of the team.

That quickly evolved into us running full-day events dedicated to MongoDB. They weren’t sales pitches, but rather technical sessions from both MongoDB engineers and community members. Soon we were running events almost every month. These have since evolved into MongoDB’s .local event series.

As I write this, I am struck by the fact that connecting with people one-to-one (sending them a coffee mug) and in-person (at meetups and conferences) were such old-school, low-tech ways of building momentum. Yet they were critical to the early adoption and enthusiasm.

Going online with MongoDB University

The one-to-one approach was critical, but it couldn’t scale. We needed a way to bring the same sense of welcome and momentum to developers we’d never meet in person.

I can’t take credit for the next major marketing innovation, which was MongoDB University. My friend Andrew joined the company with the mission of launching massive open online courses (MOOCs) that could teach thousands of developers around the world.

Developers were thrilled to post their certificate of completion of the first course, M101 MongoDB for Developers, on LinkedIn and Twitter. This small piece of recognition became a source of pride for those that took the first course. And those social media posts fed the virtuous cycle, driving more virality and momentum around the product.

BTW, the last time I checked, over a million people have taken an online course on MongoDB.

There was a lot of other stuff to

There was a lot of other stuff too. Stickers mailed to university students. Answering a post on our free support forum at midnight. Posting the logo of every company using MongoDB on the docs. None of it looked like “marketing” in the traditional sense.

I think that’s the point. The early success of MongoDB wasn’t a campaign. It wasn’t a launch moment or a viral video… though the infamous “MongoDB is Webscale” video certainly helped 🙂 It was thousands of small interactions. That accumulated into a community, and the community became the engine.

(MongoDB would get a lot of flak on Hacker News for the MongoDB is Webscale video, with many developers assuming that we made the video as a marketing stunt. Another fun fact: I still to this day have no idea who made the video, but rest assured it wasn’t anyone who worked at MongoDB. )

The playbook, if you can call it that: make the product so good it sells itself, then treat every person who shows up for it like they’re part of something. The rest tends to follow.