Sparround

Which metrics to watch, and which to ignore

Store consoles show dozens of numbers, and that can do more harm than good: you check them daily and never make a decision.

Metrics split into two groups:

  • Vanity metrics — pleasant to watch grow, but they never change a decision: total installs, total followers, total views
  • Decision metrics — when they move, what you do changes: conversion rate, activation, retention, CAC, LTV

A simple test: "if this number doubled, what would I do differently?" If there is no answer, it is a vanity metric. You may look at it, but you do not decide by it.

The most common vanity metric is total installs. It only ever goes up (never down), so it always looks like good news — even while the app is dying.

MetricWhat it tells youWhen you look at it
Store conversion rateWhat share of page viewers install2-4 weeks after an ASO change
Activation rateWhat share of installers reach the first real outcomeAfter changing onboarding
D1 / D7 / D30 retentionThe share returning after 1, 7 and 30 daysAlways — it is the product's health indicator
ChurnThe share who leave — the inverse of retentionMonthly, in subscription apps
CACWhat it costs to acquire one userWhen you run a paid channel
LTVHow much revenue one user brings over their lifetimeBefore moving to paid
Total installsAlmost nothing — it only ever growsNever, for decisions

Retention — why it is the central metric. The retention curve shows at a glance whether the app is healthy. It takes one of three shapes:

  • A curve falling to zero — users arrive and all of them leave. The app builds no habit; more marketing does not improve the situation, it just enlarges the loss
  • A flattening curve — the curve settles at some percentage (say 20%). This is a healthy sign: 20 of every 100 users stay, so as marketing grows the app grows
  • A rising curve — rare, usually in products with network effects

The most important question is: does the curve flatten? If it does not, fix the product before spending on marketing.

Which number counts as "good"? There is no universal answer — it varies enormously by category, market and app type. The "D1 retention should be 40%" figures you find online come from a market that is not yours. The practical approach: compare against your own past numbers. If this month's D7 beats last month's, you are moving forward.

CAC and LTV — the simple definition. CAC = money spent on a channel ÷ users it brought. LTV = the revenue one user brings over their lifetime. For paid advertising to make sense, LTV must exceed CAC — with margin, since store commission, tax and other costs exist too. Advertising without knowing these two numbers is like buying goods without knowing the price.

Practice. Go through every metric in your store console and write for each: "if this number doubled, what would I do differently?" Strike out the ones with no answer. Done means: you have a short list of four or five metrics to track weekly, each with its decision sentence beside it.

📚 Sources and documentation