Five stages, two of which are usually not worth instrumenting. Knowing which two saves you a month of dashboard building.
AARRR pirate metrics — acquisition, activation, retention, referral, revenue — is a useful way to locate where a business leaks. For a one-person SaaS, two of the five stages are usually not worth instrumenting, and knowing which two saves you a month of dashboard building.
Two stages instrumented well beat five instrumented badly.
AARRR is normally read left to right, which puts acquisition first. That ordering is wrong for a small product, because acquisition sends people into a leaking bucket.
Start at activation. If under a fifth of signups reach first value, every additional visitor is wasted and no amount of channel work compensates. Fix that, then retention, then acquisition — the onboarding framework covers the first and the churn framework the second.
| Stage | What to measure | Worth it? |
|---|---|---|
| Activation | Share of signups reaching first value | Yes — instrument first |
| Retention | Cohort survival at month 3 and 6 | Yes |
| Revenue | MRR and its direction | Track, do not steer by it |
| Acquisition | Qualified signups from one channel | Yes, but keep it simple |
| Referral | — | Skip unless a loop exists |
Referral is the stage most solo founders instrument prematurely. Measuring it before you have a mechanism that produces referrals gives you a number that sits at zero and tells you nothing — growth loops versus funnels covers whether a referral mechanism is available to your product at all.
The standard treatment breaks acquisition into channel-level funnels with attribution. With one channel and forty signups a month, attribution is unnecessary — you already know where they came from.
One number is enough: qualified signups per month from your channel. Add source tagging only when you genuinely run two channels, which for most solo products should be later than they think. The channel framework covers why sequential beats parallel at this scale.
Placing revenue as the final stage implies you work on it after the others. In practice revenue moves when activation, retention or price change, so there is rarely a distinct "revenue workstream" for a solo founder.
The exception is pricing, which sits outside the AARRR model entirely and is usually the single biggest change available. A price increase moves revenue without touching any of the five stages — raising prices covers the sequence.
If you have fewer than forty customers, read all of these as cohort shapes rather than percentages. One person joining or leaving moves every stage's number by several points, and reacting to that movement is how founders rebuild things that were working.
Most solo founders get more from four numbers than from five stages: activation rate, trial-to-paid conversion, monthly churn with reasons, and MRR direction. That set maps roughly onto AARRR with referral dropped and acquisition simplified.
The four metrics that matter under $10k MRR covers the short list and what to ignore, and the weekly cadence covers reviewing them without building a dashboard nobody opens.
A framework splitting a business into five stages — acquisition, activation, retention, referral and revenue — to locate where it leaks. For a solo SaaS, two of the five are usually not worth instrumenting.
Activation, not acquisition. If under a fifth of signups reach first value, every additional visitor is wasted and channel work cannot compensate. Activation is also one event to instrument, which takes an afternoon.
Usually not. Measuring referral before your product has a mechanism that produces referrals gives you a number sitting at zero. Check whether a referral or viral loop is mechanically available before instrumenting it.
One number: qualified signups per month from your channel. With one channel and forty signups, attribution is unnecessary because you already know where they came from. Add source tagging when you genuinely run two channels.
As a way of locating a leak, yes. As a set of percentages, no — one person joining or leaving moves every stage by several points. Read the stages as cohort shapes rather than rates at that scale.
Tell Marcus what you currently measure. You get the single event worth instrumenting first and what it will tell you.
Try GhostCoach free →14-day free trial · cancel anytime · 30-day money-back on Lifetime