The Science Of Scaling GTM

Everyone throws around "product-market fit" like it's a binary switch.

Even my partner Cailen says that when he’s in rooms with VCs, everyone has a different definition of it:

Mark forces you to get way more scientific about it.

His framework boils down to: P% of customers achieve E event(s) within T time.

Some real examples from companies that have nailed this:

I love this because it forces you to scientifically think about leading indicators of PMF rather than just vibes.

As you test, this changes.

Could adding users (10+) or integrations (3+) get you there? Only you can answer that for your own business. But the framework forces the rigor.

You can also pair this with the Superhuman approach (40% of surveyed customers would be "very disappointed" without the product) for a less objective but still numbers-based gut check.

Shout out to Superhuman - you’d have the rip the tool out of my cold, dead hands before I let you give that away. Speaking of, try a free month using my link here.

This is the one that a lot of AI companies probably don't want to hear right now.

Everyone is running around saying "we have PMF, we've cracked it, we've got this."

But then you look at the retention metrics and... they haven't even seen a full renewal cycle yet.

Mark is clear: the tech sector considers annual customer retention above 90% to be world class. That's PMF. Everything else before that is a leading indicator.

The problem is that retention is a lagging indicator. It takes quarters or even a year to see the real number. Which is exactly why the P, E, T framework from takeaway #1 matters so much.

Retainment numbers are just the final confirmation on whether you have it or not.

When I interviewed Liam Mulcahy on the Vibescaling Podcast, he said the exact same thing. One of the biggest mistakes is always making ICP too broad and not focused enough on messaging.

Your ICP should be specific: title, company size, geo, industry, tech stack. And it should be defined by which segment has the best retention and LTV metrics (which leads to strong retention, aka good signs of lagging PMF). Not which segment has the most inbound or the lowest CAC. That leads to a leaky bucket.

This was one of my favorite concepts in the book. PMF and GTM fit are two different things.

GTM fit = acquiring and retaining customers consistently and profitably.

So you've proven people want your product. Great. But can you actually sell it through predictable channels vs. just getting a couple million in revenue through VC intros and warm referrals? Those are two very different things.

Mark's framework has three sequential stages: Product-Market Fit (customer retention), Go-to-Market Fit (scalable unit economics), and Growth and Moat (revenue growth rate). Most companies try to jump from stage 1 to stage 3 and skip the middle entirely.

This is probably why a lot of what Mark advocates in the book won't happen for many companies. But it should.

The startup ecosystem often adopts a top-down approach to annual planning. What's the revenue target we need to hit? Cool, now how do we back into the number of reps and pipeline to get there?

VCs measure every quarter based on what aligns to their IRR and fund markups, which is not always incentivized with the truth on the field. There are misaligned incentives between what leads to sustainable long-term business health and what looks good in the next board deck.

Not right or wrong. Just the rules on the field if you want to play the VC, high-growth game.

Mark has always been vocal about this, and I've always respected it.

His mantra: establish a pace and then watch the speedometer.

The best-in-class scale plan is not 10 salespeople next month and then see what happens. It's 2 salespeople a month for the next six months. If things break, you stop and fix them. If they don't, you go faster.

I saw this play out the opposite way at Celonis (check out their bad RepVue scores here to validate it). Big round, big hiring plan, back into the number of sales hires you need. It just doesn't work that way. Most of the time, these companies hire 10 reps in January and 2 are left by December.

This is one of the biggest mistakes we see. It's also why we don't take on certain recruiting opportunities at Vibescaling. If a company wants to hire X AEs next month without the demand to support it, that's a red flag.

This one connects directly to #6.

The best way to hire great people is with a bottoms-up plan: what does the demand allow us to hire into?

Versus a top-down plan: the board wants $20M next year, so we need X reps by Q1.

This is another reason we say no to certain recruiting opportunities. When companies want more of a top-down plan, the reps they hire are set up to fail.

And the best candidates can smell that from a mile away. Top talent wants to join a company where the demand exists to support their ramp, not one that's hiring on hope. They look to have as many of these 12 traits as possible.

This is literally the thesis of the entire book. And I couldn't agree more.

Sales is a science, like finance, engineering, and marketing. It's too critical to be viewed simply as an art form.

In fact, most of this book is the numbers behind how you build and scale a revenue function. The P, E, T framework. The LTV/CAC > 3 formula. The cohort analysis. The unit economics decomposition. Mark treats every GTM decision as something that can be measured, tested, and iterated on.

The art is the cherry on top. The human connection, the storytelling, the ability to read a room. But the foundation is math.

I wrote a past article on this and genuinely believe it. The best sales orgs are built by engineers and scientists who happen to love the craft of selling.

When I was building my own startup, this was the lesson that hit hardest.

Your mom will always tell you that you have a good idea. She doesn't want to hurt your feelings and she's never going to pay for it anyway. So how do you abstract away the politeness and get to truth?

You ask if they’d pay for it.

This is why paid pilots are so important. If someone is tasked with making the actual decision to pay or not, the conversation gets to a true positive or negative fast. Free pilots let everyone be polite. Paid pilots force a real decision.

I always found The Mom Test as a great book on this + how to ask better questions to see if people will buy vs. not buy (or if they really value your solution).

Liam Mulcahy said this too on our pod. He calls it the "revenue-driving product manager."

This is why, in my opinion, bankers and consultants make amazing founding AEs. They have the analytical rigor, they can synthesize feedback, they can prioritize ruthlessly. We just need them to actually want to do sales 🙂.

The job in the beginning is to take all the feedback from customers (they'll give you 10 things) and prioritize the two or three that are going to move the needle.

Pareto, 80/20, whatever you want to call it.

Mark is clear that the first sales hire should be comfortable in ambiguity, motivated more by innovating than making money, and able to pattern recognize from customer conversations and feed that back to product and engineering.

That is not your typical coin-operated AE profile.

As we’ve said, the best AEs are not coin operated. They’re builders.

I agree with this. In the early days, it might make more sense to have a higher base, higher equity, but lower variable to align incentives with the long-term sustainability of the company.

If you're still figuring out PMF, why would you put your salesperson on a traditional revenue comp plan? You're incentivizing them to close deals fast rather than close the right deals and learn from the market.

Mark suggests considering no sales comp plan at all early on. Just base salary and equity like everyone else on the team. If PMF takes longer than expected, the salesperson shouldn't be the only one who suffers financially.

And then make sure to flip this to more variable comp once there is predictability and you've graduated to the GTM fit stage. 80/20 base-to-variable in the early days. Then shift as the playbook hardens.

I saw this at a past company. A handful of off-ICP logos started dictating the product roadmap. We were building features we couldn't repurpose, couldn't build upon, and couldn't productize for the broader market.

It feels good in the beginning. Revenue is revenue. But then those off-ICP customers start churning because the product was never really built for them. Or worse, they stick around and pull your roadmap into a spiral that alienates your actual target market.

Mark frames this as a customer retention problem that originates in sales. If your reps are incentivized to close anything that moves, they will. And then CS inherits the mess.

This one connects directly to #12. If the problem is reps selling bad deals, the fix is in how you comp them.

Mark advocates paying 50% of the commission when the buyer purchases and the other 50% when the buyer achieves the leading indicator of customer retention. That's a fundamentally different incentive structure than "close the deal, get your check, move on."

Stripe did something similar. They comped reps on usage in year one. This is common for consumption-based businesses. But even for non-consumption businesses on license or enterprise contracts, you could structure it so the rep sees the rest of their commission when the customer renews.

Smart. It incentivizes reps not to sell bad deals.

You could also put a kicker when there's a renewal. Good for retention, good for NRR, and it keeps the rep invested in the account long after the ink dries.

This is where I want to add my own take on top of Mark's framework.

Mark talks a lot about "film reviews" where the team listens to recorded sales calls together and coaches off of them. He's right. This is one of the most effective ways to level up a sales team and codify the playbook.

The biggest issue? VPs don't have time. They're pulled in 15 directions and coaching falls to the bottom of the priority list. But it's arguably the highest-leverage activity a sales leader can do.

This is where AI is starting to close the gap in two ways:

The combination of these two things means you can build a coaching culture even when your leadership team is stretched thin.