Draft — under review. Last updated 19 October 2026. Where a fact is still open, this page says so rather than guessing.
Use case

Find the acquisition channels that bring retained users

Which channels bring people who stay? Answering it well means comparing cohorts that have had enough time, showing how big each one is, and saying plainly what the data can't tell you.

Who does it

A growth lead or marketing owner, with an analyst agreeing the retention definition.

What data it needs

  • An acquisition channel on each user, captured at signup or attributed later.
  • The repeat-behaviour events that define "retained" — for example, an order in week 4.
  • Enough elapsed time: a cohort that signed up three weeks ago has no week-4 retention yet.
  • Spend per channel, if the question is about profitability rather than retention.

Step by step

StepStatus
Agree the retention definition and record itLive retention views · In build catalog approval
Build weekly cohorts by signup week, split by channelLive cohorts and retention
Exclude cohorts too young to have reached the measured weekLive — shown as empty cells
Ask the question in plain language, from the product or an AI clientIn build Ask · Planned MCP

What the answer includes

The retention definition, the cohorts included and their sizes, and both the retention rate and the count of retained users. A small channel can have the highest rate while a large one brings the most retained users, so the two are shown separately.

If spend per channel isn't in the catalog, the answer says it can't tell you which channel pays back — rather than implying it.

The decision it supports

Which channel to look into, or which to give a controlled budget test. Not an automatic reallocation of spend on retention alone.

Questions this page doesn't answer are the ones we want to hear. Ask on a call, or write to contact@depsmith.com.

Request a demo