A framework by Jason Lemkin
Jason Lemkin's ICP-and-churn diagnosis reads a high churn rate before you blame the product. Most SaaS churn is a customer problem. When churn runs high for your segment, you sold to the wrong ideal customer profile. The product is seldom the weak part. Fix the buyer and the retention math fixes itself.
The diagnosis in one line
You have something, but it ain't SaaS.
Lemkin's line to founders at a million in revenue with abnormal churn. The product can be good. The segment is wrong.
The framework
When churn climbs, the instinct is to look at the product. Add the missing feature. Fix onboarding. Ship faster. Lemkin's diagnosis starts somewhere else. Before you treat churn as a product gap, ask whether the number is normal for the customer you signed. A 3 to 4 percent monthly churn rate is endemic to small businesses. The same number from an enterprise book of business signals a serious problem. The rate only means something relative to the segment, and most teams never set that baseline. They benchmark against a number from a blog post instead of against the buyer sitting in their own pipeline.
The reframe matters because the two diagnoses lead to opposite work. If churn is a product problem, you build. If churn is a customer problem, you change who you sell to. That is positioning and go-to-market, not roadmap. One is a quarter of engineering. The other is a different sales motion. Sometimes it means a different price point or a different buyer. Lemkin's pattern from years of investing is that abnormal churn almost always traces to the second cause. Small businesses cancel everything the day the spend stops paying back. That is not a bug you can patch. It is the economics of the buyer you chose.
So the framework is a triage, not a fix. Establish what churn is normal for your ICP. Compare your actual number to that baseline. Segment the cohorts so the blended average cannot lie to you. Only then decide whether you are looking at a weak product or a wrong customer. Most of the time, Lemkin's view is that you are looking at the wrong customer wearing a product costume.
How to apply it
This is a sequence of honest questions about your buyer, not a churn-reduction checklist.
Decide what monthly churn is normal for the customer you actually sell, before you judge your own. Very small businesses churn 3 to 4 percent a month as a rule; mid-market and enterprise should be a fraction of that. The baseline is the segment, not an industry average.
If you sell to enterprise and lose 4 percent a month, that is abnormal and almost certainly a customer-fit problem. If you sell to very small businesses and lose 4 percent, that is endemic, and the question shifts to whether your model can survive it.
Ask whether churned customers left for a competitor or left the category entirely. Leaving the category is a segment-fit signal, not a feature-gap signal. No feature wins back a buyer who never needed the category.
SMB revenue grows fast early because you acquire customers in a day, not a year, which masks a churn rate compounding underneath. Look at retained revenue cohorts, not new logos.
Normal churn for the segment with the product as the constraint means build. Abnormal churn for the segment means the work is changing who you target.
SMB software has to be excellent, self-serve, and product-led, because the customer can and will cancel the moment it stops paying back. That is a requirement of the segment, not a strategy you opt into.
As you move up-market or add a segment, the acceptable baseline moves with it. Treat the diagnosis as a recurring read, not a one-time audit.
On what he does when he sees a number that is abnormal for the segment, Lemkin was blunt.
I have passed on every single company that had, for their segment, abnormally high churn. I've gone back to so many founders that were at a million with 78% churn, I'm like, you have something, but it ain't SaaS. It's not SaaS.
Notice he does not say the product is bad. He says it is not SaaS. The business may be real. It is just not a high-retention software business, because the buyer it attracted does not behave like one. That is the whole diagnosis in two sentences.
Boundary conditions
The diagnosis is not a license to stop building. Lemkin is explicit that the hard segment has its own non-negotiable bar:
They'll look at their credit card statement and they'll cancel everything. It's brutal. So SMB software has to be better, it has to be self-serve, it has to be PLG. It's not a choice, it's a requirement.
The honest caveat: this is a triage tool, not an excuse. If the answer is wrong customer, the fix is positioning and go-to-market. If it is wrong product for a segment you have chosen to keep, you still have to build. For SMB the bar is higher, not lower.
The sources
20VC · E1157, SMB churn segment
Why the rate only means something by segment
If the churn is anything more than a 3 or 4% a month it's not even software anymore. It's some sort of consumer-like thing that does not have the same retention.
20VC · E1157, SMB churn segment
Passing on abnormal-for-segment churn
I have passed on every single company that had, for their segment, abnormally high churn. You have something, but it ain't SaaS.
20VC · E1157, SMB churn segment
The SMB requirement
They'll cancel everything the day that doesn't make the money. So SMB software has to be better, self-serve, PLG. It's not a choice, it's a requirement.
All three excerpts come from the same 20VC episode (E1157, the SMB churn segment). Listen to the full passage from 53:00 onward for the complete argument.
Where experts disagree
Jason Lemkin
says churn that is abnormal for your segment almost always means you sold to the wrong ideal customer profile. The product can be fine and the buyer wrong, so the fix is positioning and go-to-market, not a quarter of engineering.
Bob Moesta
says the customer hired you for a job, then their context changed and the job moved on, so the switch that brought them in reversed. The fix is to interview the churned and find the struggling moment, not to re-segment.
Alex Hormozi makes it three ways: even the right customer with the right job leaves if they never reached the result, which puts the fix in the first thirty days. The three diagnoses lead to completely different work, so run all three before you commit.
Want the full playbook?
From Hamilton Helmer, Bill Carr, Anuj Rathi, and 48 more. Drop one .md into Claude, Cursor, or ChatGPT. Your AI cites practitioners, not guesses.
Instant .md download · One-time purchase · No subscription
Related frameworks
Retention
Churn Is a Delivery Problem
Alex Hormozi
People don't churn because the product is wrong. They churn because they never got the result. Fix onboarding, expectations, and demonstrated value.
Research
The Opportunity Solution Tree
Teresa Torres
One outcome at the root, unmet needs below it, solutions and assumption tests below those. Prune the rest.
Pricing
The Value Metric
Patrick Campbell
Pick the unit you charge on before the number. Get it right and expansion revenue typically doubles.
Research
The Reference Customer Method
Christian Idiodi
Product-market fit is not a score. It is a count of people who will put their reputation behind you.
