Segment Your SaaS Data:The Average Is a Fiction

Segment your SaaS data: your aggregate SaaS metrics are a fiction. The same product can churn 2.1% of newsletter-sourced customers a month, 8.7% of Product Hunt customers, and 1.4% of referred customers — while the blended average reads 4.2% and tells you nothing. Segmenting your data by acquisition source is how you find out where your growth actually comes from, and where your effort is being wasted.

Here's the segmentation exercise, a real cohort breakdown, and how to turn what it reveals into your growth levers.

Segment your SaaS data: why the average is a fiction

Most founders never find their growth levers for one reason: they don't segment the data; they look at averages. And the average is a fiction. Inside every average there are a few customers who are amazing and a long tail who are mediocre. The amazing ones are the lever. The average blends them together until neither is visible.

This matters because, after three to six months of running your five-metric dashboard, most of your growth turns out to come from one or two specific things. One acquisition channel that converts at three times the rate of the others. One feature that activated customers use ten times more than the average. One pricing tier with half the churn of the others. One referral source that produces customers who themselves refer.

You cannot see any of that in an aggregate number. You can only see it by breaking the number apart.

A real segmentation: same product, three customer profiles

Pick any one of your metrics — say, monthly churn — and instead of looking at the overall rate, break it down by source. Here's what that looks like in practice:

Same product. Same business. Three completely different customer profiles depending on how they got in. The newsletter and referral cohorts compound; the Product Hunt cohort bleeds money, because acquisition cost is high and lifetime is short. And the aggregate average — 4.2% monthly churn — hides this story entirely. It's not wrong, exactly. It's just useless.

What the segments tell you to do

The growth lever in this example is obvious from the segmentation and invisible without it: stop putting effort into Product Hunt-style amplification and pour it into newsletter growth and referral mechanics, because customers from those sources are five times more valuable over their lifetime than the Product Hunt customers.

Notice what just happened. A question that founders normally answer with vibes — "where should my marketing effort go?" — became an arithmetic problem with an unambiguous answer. That's the entire promise of growing on data instead of feelings: the segmented view converts strategy debates into subtraction.

It also settles the launch-platform anxiety most solo founders carry. A big launch spike feels like growth. But if the cohort it brings churns out in four months at €42 a head, the spike was theatre. The unglamorous channels — a newsletter you grow subscriber by subscriber, customers referring customers — are where the compounding lives.

12 weeksYour corporate career was never the planApply to the Solopreneur Accelerator.Apply now →

Finding your growth levers

Growth levers are the few inputs that, if you double down on them, produce disproportionate output. Most businesses have one or two real levers and a dozen distractions that look like levers. The job, by month six post-launch, is to know what your two levers are and to put eighty percent of your remaining effort on them.

How do you find them? You look across the data and ask, for each metric: which segment of customers is best? Which channel? Which feature? Which timing? The segments that are unusually good are pointing at the lever. Then the work inverts — figure out how to acquire more customers like the good ones, route more traffic through the high-converting channel, build more features that resemble the high-activation one.

Find the amazing customers. Study them. Optimize for them. The mediocre tail will take care of itself — or churn, which at this stage amounts to the same thing.

The monthly segmentation hour

Churn by source is one cut. The same exercise applied to activation rate, to MRR-per-customer, to support ticket volume, to feature usage — each one reveals patterns the average obscures. Which cohort activates fastest? Which segment files the most tickets relative to what it pays? Which feature do your longest-lived customers use that everyone else ignores?

Spend one hour each month on segmentation analysis. That's the whole practice. The patterns it reveals are usually the most important thing you'll learn that month about your business — and each pattern converts directly into a hypothesis you can feed into your weekly growth experiments. Segmentation finds the lever; experiments confirm it moves.

The bookThe Solopreneur RevolutionLuis Gonçalves on building a solopreneur business with AI while employed.Get the book →

Frequently asked questions

How do I segment my SaaS customers?

Start with acquisition source — it's the segmentation that most often changes decisions. Tag every customer with how they arrived (newsletter, launch platform, referral, organic), then compare churn, lifetime, and MRR per customer across the tags. One cut by source is worth more than ten demographic cuts.

What is a growth lever?

One of the few inputs that produces disproportionate output when you double down on it — a channel converting at three times the rate of the others, a feature activated customers use ten times more, a pricing tier with half the churn. Most businesses have one or two real levers and a dozen distractions that look like levers.

How often should I run segmentation analysis?

One hour a month is enough for a solo founder. Weekly you watch the five dashboard metrics; monthly you break them apart by segment. The monthly hour usually produces the most important insight you'll get that month.

Why do my Product Hunt customers churn so fast?

Launch-platform cohorts are browsers, not buyers — they arrived for novelty, not for your problem. In the example above they churned at 8.7% monthly with a 4-month lifetime versus 1.4% and 24 months for referred customers. That's not a product failure; it's a signal about where your durable customers actually come from.