Marketing experiment tracker

Keep hypotheses, results and decisions in a reusable learning library.

Replaces: the experiment log in Google Sheets

Shiftline · Marketing experiment tracker

Marketing experiment trackerShiftline

Tessa GrantGrowth marketing leadTG

Running: 5; Collecting results
Concluded: 18; Since December 2025
Wins: 7; Adopted after review

Primary metric	Result	Review date	Status
Trial signup rate	Day 9 of 14	Nov 19	Running
Day-7 activation	+12% activation	Nov 14	Review due
Cost per demo	$412 vs $250 cap	Nov 20	Guardrail hit
Paid conversion	Too few signups	Nov 13	Inconclusive
Click to trial rate	Day 4 of 21	Dec 1	Running
Demo form completion	+14% signups	Nov 10	Win
Trial signup rate	No difference	Oct 26	No change
Trial signup rate	+9% signups	Oct 12	Win

Showing 8 of 23 experiments

Ran today at 7:00 AM

Refreshed results for 5 running experiments from PostHog and Google Analytics

Onboarding email: video first

Review date today · 2,860 trials

Review due

Decision note, drafted for review

Onboarding email: video first. Recommend: adopt variant B. Result snapshot from Nov 14: 38.2% of trials who got the setup video published a schedule within 7 days, against 34.1% for the text checklist (+12%). PostHog puts the probability that B is better at 97%. The test reached 2,860 trials against a plan of 2,800, at the planned review date. Guardrail: unsubscribes were 0.6% vs 0.5%, under the 1% limit. Follow-up: make the video email the default for all new trials and recheck day-7 activation in 4 weeks. Owner: Tessa Grant.

Hypothesis: A 2-minute setup video in the first email gets more trials to publish a schedule by day 7
Variants: A: text checklist · B: setup video
Primary metric: Trials that publish a schedule within 7 days
Result snapshot: Nov 14: B 38.2% vs A 34.1% (+12%), PostHog probability B is better 97%
Sample: 2,860 trials; the plan called for 2,800
Guardrail: Unsubscribes 0.6% vs 0.5%, under the 1% limit

History

Nov 14Planned review date reached; decision note drafted
Oct 17Experiment started; plan v1 locked

Runs on its own

Refreshes results and flags review dates

Each morning it refreshes results from PostHog and Google Analytics, snapshots them with sample coverage, and flags review dates and guardrail breaches.

Waits for you

Calling a winner

Calling a winner is yours. It won't show a result before the planned end date, and every decision links to the snapshot it was based on.

Works with

PostHog
Google Analytics
Google Sheets

What it keeps track of

Experiment: Hypothesisaudienceownerreview date
Plan: Versionvariantsprimary metricguardrails
Result snapshot: Datevaluessample coveragesource
Decision: Adoptstop or continuerationalefollow-ups

Make it yours

More apps like this

All 42 apps