How to run a growth audit: four vitals, in this order
Every engagement starts with a channel question. "Should we scale Meta?" "Is TikTok dead?" "Why isn't this funnel converting?"
Channel is the last thing I look at.
A doctor does not average your cholesterol with your blood pressure into one health score. Four vitals, four separate lines, four separate pass or fails. A growth audit should work the same way. A blended score hides which system is failing. Four separate lines do not let it hide.
The order
- Positioning. A channel cannot fix a story nobody believes.
- Retention. Acquiring users into a leaky bucket just raises the price of the leak.
- Unit economics. A channel that works on paper can still lose money per user.
- Channel. Last, on purpose, every single time.
Vital 1: Positioning
There is no number here. No spreadsheet measures whether a stranger repeats your pitch back correctly. The test is qualitative: show your store page or landing page to someone in your target audience and ask what the product does and who it is for. If they cannot say it back, every channel will pay more than it should.
Vital 2: Retention
Pass: the curve flattens by day 30. Fail: it is still falling at day 60 and day 90.
There is no fixed floor. A hardcore RPG and a casual match-3 game do not share one. The test is the shape, not the level: does the bleeding stop by day 30, or keep going?
A mobile-gaming studio passed every chart I ran in the first week and failed this one quietly for two months. We fixed the leak, not the spend. Average daily revenue went from $3,524 to $4,081, up 15.8%.
Vital 3: Unit economics
Score it by payback day:
| Breakeven day | Verdict |
|---|---|
| By day 90 | Scale-ready |
| Day 91 to 180 | Watch the payback closely |
| Later, or never | Underwater. Every dollar of growth is renting the illusion. |
Vital 4: Channel
Not a fixed CPI target. Score each channel against the CAC ceiling that vital 3 just set. The exact same CPI passes for one business and fails for another.
Channel goes last because it is the only vital everyone wants to check first.
What skipping the order costs
A wellness subscription app once asked me to scale its paid spend. The diagnostic never got to channel. The bigger lever was in the product, not the budget. We added product-led growth: referrals and reward-based features. Daily revenue grew 23% month over month.
Skip the order and the channel question still gets answered. Just fast and wrong. A bigger budget on a product that cannot grow itself just raises the cost of every user.
The move
- Score each vital on its own line. Never combine them into one health score.
- Go in order and stop at the first failure. That failure is your constraint.
- Fix the constraint before you touch the next vital.
- Count the failures. One failure is a project. Two or more is a leadership gap, not a staffing gap.
Bottom line
Diagnose in that order, or you are buying an expensive fix for the wrong problem.
FAQ
What should a growth audit include?
Four separate checks in a fixed order: positioning, retention, unit economics and channel. Each one gets its own pass or fail, so a strong channel cannot hide a failing retention curve.
Why check the channel last?
Because channel performance depends on the other three. A channel cannot fix weak positioning, a leaking retention curve or a payback window that never closes. Checking channel first usually leads to spending more on the wrong problem.
What is a good retention benchmark for an app?
There is no single benchmark that fits every product. The more useful test is the shape of your own curve: it should flatten by around day 30. If it is still falling at day 60 and day 90, retention is the constraint.
What payback period is good for a consumer app?
Breakeven by day 90 is a strong position to scale from. Day 91 to 180 can work but needs close watching. Later than that, or never, means growth is being bought at a loss.