Channel Adoption Readout

Synthetic data

Which channels and pages brought people who kept using the product, not just visitors.

The subject is made up: a free-to-start, open-source workflow product with a cloud signup and a demo-request form. Weeks 1-8, no calendar dates. Seed 20261002.

Loading the readout.

Sources

Every source, with each signup credited once. Pick a row to see its weeks and the paths people took before signing up.

Credit
Activated means
in the first 7 days

Attention or adoption

Each dot is a source. Across is its share of all visits, up is its share of all activated signups. A dot above the line brought more than its share of people who kept using the product.

Adoption Mixed (below the line, within 2x) Attention only (2x or more below) Too early to judge

Both axes use a square-root scale so the small sources don't pile up in the corner. The dashed line is still where the two shares are equal. No source has no visits, so it isn't plotted.

Trace the spike

Signups per week. Pick a week to see where its signups came from and how many of them activated.

Activated Not activated Held burst

Source Signups Share Activated Rate

Exceptions

Everything the rules set aside, so nothing disappears quietly.

Not counted above

Open-source activity is real interest, but it can't be tied to a person or a signup, so it gets its own panel and is never added to any number above.

The rules

Every number on this page comes from these eight rules, computed live. The grey line under each one is its count in the current view.

    Assumptions

      Questions I'd want answered before trusting this on real data

      1. What event counts as an activated user for you today, and how long after signup do you check?
      2. Do self-hosted installs send anything back, or is that usage invisible to marketing?
      3. When someone already on the free plan requests a demo, does the credit go to the page they first read or to the product?
      4. When someone reads three pages before signing up, which one does your reporting credit today?
      Next: what to make and test next →