The audience that was recomputed later
Six weeks on, the query returns a different list, and the report describes people who never saw it.
Plan the motion, freeze who it went to, read the result against it.
A campaign records a hypothesis, audience criteria and a channel. Freezing writes a recipient snapshot: who matched, who was excluded and under which rule, with the date supplied by the server. Afterwards you write what was measured, and the studio reads that against the hypothesis. Nothing is sent from Growth-OS at any point.
In the app
Captures of Growth-OS itself, rendered with sample data rather than our own records.




What goes wrong when this work lives in spreadsheets, documents and one person's memory.
Six weeks on, the query returns a different list, and the report describes people who never saw it.
An unreadable clause refuses the freeze, and up to five of them are reported with a count, instead of the campaign going out to everybody.
Without a written hypothesis, every outcome is a success. The studio reads the outcome against what you said would happen.
Highlights
A live preview keeps moving until you freeze it. After that the snapshot is the campaign audience, and it is never replaced.
Seven rules in order, and exactly one is recorded per account: the first that matched. The full list travels with the record.
An angle derived from evidence records shows it. One nobody can support is badged as a hunch, with no record id to render.
From an empty record to something you can act on, in the order you do it.
Write the campaign
Hypothesis, audience criteria and channel. The criteria use the same clause grammar as a segment.
Read the studio
Angles per family, and a live recipient preview with the counts still moving.
Freeze the snapshot
The server supplies the date and writes one snapshot linked to the campaign. A second freeze is refused.
Run it elsewhere, record it here
You take the motion to the channel yourself. Then write the measured outcome and set the status to done.
Honest limits
Nothing is sent. No mail provider, list tool or sending API is contacted anywhere in this feature, and a snapshot records who would have been contacted rather than who was.
The measured outcome is typed by a person. One campaign with no comparison cannot show that the angle caused the result.
These pages share records with this one, so what you write here shows up there.
The page opens on this project's records. The in-app guide describes every field.