Laws of PM Growth Retention is the Cake

Retention is the Cake

Acquisition Is the Icing

Acquisition is the icing. If the cake is dry, more icing just melts off faster.

Why PMs should care

A product with good acquisition and bad retention is a leaking bucket. Every pound you spend on marketing runs straight out the bottom, and the only thing that changes as you scale spend is how fast the leak grows.

Retention work is almost always harder, slower, and less visible than acquisition work. There's no satisfying weekly chart that ticks up as you ship retention fixes, because the payoff is a cohort curve that flattens out twelve weeks later.

But no acquisition channel scales forever without the underlying economics working. Companies that confuse headline growth with actual growth usually find this out in the quarter their cost-per-acquired-user doubles and their lifetime-value per user hasn't moved.

Rule of thumb: don't scale paid acquisition until 3-month retention is at or above your sector benchmark, and never let the acquisition team grow bigger than the retention team.

Example in product work

A growth team scales paid acquisition in a quarter, hitting the 'new users acquired' KPI with room to spare.

The Q2 leadership deck looks fantastic: user growth up 180%, CAC down 12%, everything green.

Three months later, the cohort charts come in. M1 retention for the new cohorts: unchanged. M3 retention: slightly worse (the new users came from lower-intent channels). Blended LTV:CAC — the number that actually matters — now 1.4, down from 2.1.

The company has spent £4m to acquire users who will not, in aggregate, pay back their acquisition cost. The KPI was the wrong cake. The right KPI was 'M3-retained users from new cohorts', which would have turned off the spend before it did damage.

What to do when you see it

Sources & further reading

Back to all 59 laws