Start Here / Step 7: Track, Improve & Scale

Seller Ad Testing Tracker

Record a one-variable test with a baseline, maturity, cap, outcome, and explicit decision.

Requires: Seller Campaign Economics Calculator Risk limits for initial setup; 90-Day Seller Campaign Tracker Actual baseline and Weekly Campaign Review Worksheet hypothesis when evaluating performance

When to use: Before every performance change

Estimated time: 10 minutes per setup and review

Excel workbook

Two tabs

Tests: hypothesis, dates, offer/hook/copy/headline/creative IDs, changed variable, fixed variables, baseline, sample requirement, success threshold, maturation, spend cap, exposure, outcomes, calculated CTR/CPL/cohort rates, result, and Keep/Iterate/Stop/Inconclusive. Guide: column meanings, actual-versus-planning distinction, and an illustrative test description. Production rows are empty; the example is not actual campaign performance.

Fill a row before launch

Choose one metric and its denominator. Write a baseline using a defined period and mature cohort. Describe the practical effect that would justify a change, the sample/exposure you need, outcome maturation period, review date, and maximum test spend. A sample requirement is your predeclared planning rule, not automatically statistical significance.

During and after the test

Record actual spend, impressions, link clicks, raw leads, qualified, first booked, and first held from the same test-attributed cohort. Keep its ad ID and version stable. Do not use this week's bookings divided by this week's new leads. Do not count reschedules as new bookings. At review, compare counts, rates, quality, economic safeguards, and maturity. The automatic spend/sample prompts are operational checks; they do not determine a winner. Record the result and decision yourself, with the evidence and limitations.

Example test brief

Test T001, planning illustration: Keep checklist offer, area, page, budget, and nurture fixed. Change only the ad's opening sentence. Use qualified opportunity cost as the primary metric and first-held quality as the downstream safeguard. Choose a meaningful dollar/rate threshold, sample, and cap before launch. Mark Inconclusive if the exposure is too small or the outcome has not matured. Unequal Meta ad delivery limits causal claims.

You’re done when

The test has a stable ID, one variable, outcome definition, maturity window, practical threshold, cap, and recorded decision.

Return to Step 7 · Start Here

Support reference R18