Is retention more important than acquisition?

Arithmetically yes, because retention improvements apply to every user you will ever acquire. The exception is early on, when you lack the volume to measure retention at all and acquisition is the only way to get it.

What is a good day 7 retention rate?

It varies enormously by category, so compare against your own trend rather than a published figure. A collapse between day 1 and day 7 is meaningful regardless of the absolute number: it means the app was understood and not needed.

Why does one traffic source retain better than another?

Usually genuine audience differences. Occasionally it is an artefact: if a source loses taps before the store, only the persistent users arrive, and that self-selected cohort retains better while converting worse. Check click to install rate before drawing conclusions.

How do I improve retention?

Shorten time to first value. Almost every retention improvement in practice comes from getting a user to the moment the app is useful faster, rather than from notifications or rewards.

Retention versus acquisition

Retention compounds and acquisition does not, which is why the advice is always to fix retention first. It is also wrong for a specific and common case.

9 August 2026 · PAD team

The arithmetic

A ten percent improvement in day thirty retention raises lifetime value for every user you will ever acquire, past and future. A ten percent improvement in cost per install raises the value of next month's spend only. The first compounds, the second does not.

That is the standard argument and it is correct, with one exception that catches a lot of early teams.

When to fix acquisition first

If you do not have enough users to measure retention, retention work is guesswork. Below a few hundred installs a week, retention curves are noise, and optimising against noise produces confident wrong decisions.

Situation	Work on
Fewer than a few hundred installs a week	acquisition, to get readable data
Day 1 retention below 25%	onboarding, urgently
Day 1 fine, day 7 collapses	the core value moment, not the funnel
Retention healthy, growth flat	acquisition, confidently

The third row is the interesting one

Good day one and poor day seven means people understood the app and did not find a reason to return. That is a product problem wearing marketing clothing, and no amount of acquisition spend fixes it. It is also the most expensive pattern to misdiagnose, because acquisition can hide it for months.

The measurement problem underneath both

Retention is measured per cohort, and cohorts are defined by source. If one source loses a share of its taps before the store, its cohort is smaller than you think and self-selected: the people who arrived were the persistent ones who tapped twice.

Measured on 21,000 taps routed between 3 and 9 August 2026, 19,381 came from inside social apps and 95.6% were iOS. A source with a delivery problem produces a cohort that retains unusually well and converts unusually badly, which reads as a puzzling audience rather than a mechanical artefact.

What to actually track

Day 1, 7 and 30 retention by source, not blended.
Cost per retained user rather than cost per install.
Click to install rate by source, which catches the artefact above.
Time to first value, which predicts day 7 better than anything else you can measure on day one.

A reasonable rule

Spend on acquisition until you have enough volume to read retention. Then spend on retention until day seven stops improving. Then return to acquisition with economics you can trust. Most teams do the first and third and skip the middle.

A cohort that never fully arrived will mislead every retention number you compute.

Create a link