Next Tests

Synthetic data

What to make and test next, with the bar set before each test starts.

Same made-up product and the same data as the readout: Weeks 1-8, seed 20261002, first touch credit, activated = 3 or more runs in the first 7 days.

Loading the plan.

Which topics' readers kept using it

Signups credited to the integration how-to pages, the YouTube walkthrough and the docs quickstart, split by the topic each person read. Readers are visits to those pages on that topic. Sorted by activated per 1,000 readers. The held burst is left out.

# Topic Readers Signups Activated Activated per 1,000 readers Qualified demos Most common first workflow

Three tests, with the bar set first

Each test gets one metric, a minimum sample and a win bar before it starts, so the result can't be argued into a win afterwards.

Weekly check

The four numbers to review each week for each live test, filled with synthetic week 1 values. The status line applies the rules above.

Test Visits Signups Activated Qualified demos Activated per 1,000 visits Status after week 1

Assumptions

    Two more questions before running these

    1. Who records demos today, and which one led to the most real usage?
    2. Do you already use a minimum sample before calling a channel test?
    ← Back to the readout