Growth is not a religion.
Most growth plans run on hope, habit, or magical math. George Storm on cutting two thirds of a revenue org, and why a true operator acts on data instead of a story.
Part three of three: the read, the loop, and the move.
Act I. The Scripture
For a decade, SaaS growth had its own scripture. One AE closes 10k in new ARR a quarter. Add nine more, and you've got 100k. Add ten more after that, and the model says you've got a machine. Boards funded it on faith. Investors underwrote it without asking whether the tenth AE was selling into the same finite pool of demand as the first, or a fresh one. Nobody tested it. You don't test scripture. You follow it.
Growth is not a religion.
Most growth plans are still run on one of three articles of faith.
Hope, that the pipeline turns before the quarter ends. Habit, doing it this way because that's how it's always been done, not because anyone checked whether it still works. Magical math, ten reps because one rep worked, scaled up with no more scrutiny than that.
None of the three is data. All three are guesses, presented as a game plan.
That guess wasn't structurally wrong so much as regime-dependent, and the regime it depended on was cheap capital. For a long stretch of the last decade, capital was cheap enough that being wrong about the arithmetic didn't show up on anyone's P&L before the next raise closed. When the cost of capital rose and growth-at-any-cost stopped being fundable, the gap between the scripture and reality didn't close quietly. It closed as a correction.
Layoffs.fyi tracked more than 164,000 tech layoffs in 2022, then over 262,000 in 2023, the worst year on record. Public SaaS multiples fell from roughly 17x forward revenue at the 2021 peak to 5 or 6 times revenue, per KeyBanc Capital Markets and Sapphire Ventures' own survey of the market. This was companies discovering that what they'd scaled on had never been unit economics. It had been a faith capital was willing to fund, and when the regime changed, the faith stopped being enough.
Act II. The Cut
For a while, I ran a revenue team larger than it needed to be, because that size felt like the safe, defensible choice. More reps felt like more coverage. Cutting felt like admitting something. So I didn't, not until the data made the question impossible to avoid any longer: what was all that headcount actually protecting?
Last year I cut two-thirds of that revenue org.
Not because I'd found a clever new methodology. Because I finally asked the question I should have asked sooner, and the honest answer was nothing. The extra headcount wasn't protecting revenue. It was diluting the attention, the coaching, and the deal-level scrutiny the remaining reps needed to actually perform. What happened next is the part I actually want people to sit with: a third of the team, carrying the full weight of the number, immediately overachieved quota. Then overachieved the following quarter again, against a quota I'd raised, not lowered. Then again the quarter after that. Win rate doubled. Average contract value doubled.
Growth is not a religion. I didn't learn that from a framework. I learned it from watching a third of a team outperform the rest of it, at a higher bar, the moment I stopped protecting headcount out of habit and started trusting what the data was already telling me.
Act III. The Discipline
The mistake would be reading that story as "cut your team and everything gets better." It isn't a headcount story. It's what happens once you stop letting hope or habit answer a question the data was always able to answer instead, and headcount is simply the input most leaders are least willing to point that question at, because it's made of people, and because "we've always had this role" carries a weight that "we've always used this tool" doesn't.
But the same discipline has to apply everywhere, or it isn't a discipline. If the data says a role, a tool, or a process isn't earning its place, the response has to be the same regardless of how long it's been around.
Most orgs track pipeline coverage, win rate, and ramp time obsessively. Almost none of them track the efficiency of the org itself as a living number. It's not just my number to make that point with. ICONIQ's own 2025 benchmark study put median ARR per employee at 182 thousand dollars five years ago and 237 thousand now, real, industry-wide movement toward doing more with fewer people. Most companies still only pull that number out once a year for a board deck, treat it as a curiosity, then forget it until the next one. It deserves to be read the way any other signal is read, continuously, as an input to a decision, not a retrospective footnote.
And the correction never fully reversed. The Bessemer Cloud Index, the closest thing public SaaS has to a scoreboard, was still sitting around 8x revenue as of late summer 2026, nowhere near the 17x multiples of 2021. The market re-priced growth once and hasn't priced it back up. Anyone still running last decade's scripture is underwriting a regime that no longer exists.
None of this needed AI to be true. My own cut wasn't an AI decision. But it's worth being honest about why this kind of scrutiny is cheaper to act on right now than it's ever been. Removing a role has always cost more than the salary line suggests. The visible cost is compensation.
The cost nobody prices in is overhead: the hours a manager spends coordinating, coaching, and course-correcting work that a smaller, sharper unit could carry with less oversight, and the opportunity cost of what that manager could have been doing instead. AI changes that math because it absorbs enough coordination overhead that keeping a role no longer carries the same hidden management tax it used to. My team stays intentionally small even as our targets grow. We always evaluate automation or delegation before opening a new hire. The new hire rarely wins that race.
The Move
A forecast tells you what's coming. A plan tells you what to do about it. Neither one moves a company an inch. At some point the data stops being analysis and becomes a decision, and somebody has to make it. Not the model. Not the board deck. A CRO, a CEO, an operating partner, someone with their name on the outcome, choosing to act on what the data says instead of what's comfortable, and owning it whether it turns out right or wrong. That's what separates a true operator from someone who just runs a good process.
Hope was never a forecast. A forecast was never a plan. And growth was never a religion. Each one is the same mistake: mistaking a story for a signal, right up until someone with the nerve to act on the data instead of the story moves the company forward anyway.