Ask Efri

How do I know which part of my launch actually drove sales?

Efri · 1 min read
In short
  1. 1
    No single tool can tell you. You need multiple methods combined.
  2. 2
    Tag every link with UTMs. Ask "Where did you hear about us?" at signup.
  3. 3
    Platform dashboards overstate their role. Last-click understates upper funnel.
  4. 4
    For true causation, run a holdout test. Hold back a segment and measure the difference.
Full answer

Attribution is not exact. But you can get close enough to make good decisions.

  1. 1
    1. Tag every link with UTMs
  2. 2
    2. Add a self-reported attribution question
  3. 3
    3. Use a multi-touch model
  4. 4
    4. Look for patterns across cohorts
  5. 5
    5. For causation, run a holdout test

1. Tag every link with UTMs. Use utm_source, utm_medium, and utm_campaign on every link. Without them, traffic shows as "direct" and you are blind.

2. Add a self-reported attribution question. At signup or purchase, ask: "Where did you first hear about us?" This captures dark social: DMs, Slack, WhatsApp, private communities that analytics cannot see.

3. Use a multi-touch model. Platform dashboards overstate their own role. Last-click understates upper-funnel content. First-touch shows what created awareness. Linear distributes credit evenly. Pick one that matches your sales cycle.

4. Look for patterns across cohorts. Do LinkedIn-first customers close faster? Do Instagram visitors join the list and convert later via search? Which content topic appears most in buyer journeys?

5. For causation, run a holdout test. Attribution shows correlation. Incrementality shows causation. Hold back a segment of your audience or geography from one channel. Compare conversion to the exposed group. The difference is the real lift.