UGC Handbook / Lesson 14 of 14 / Chapter IV: Measure & scale

The 30-minute weekly UGC review

Lessons 01 to 13 gave you a batch, a link per video, a hook formula, creators, a pay structure, and a boost rule. None of it compounds unless something forces you to look at the numbers on a fixed day and decide.

A user acquisition manager who says they have run campaigns for more than 125 apps over 7 years wrote the sentence this lesson is built on, in a RevenueCat post on running UA as a one-person team: "Consistency will always beat intensity." The next line: checking campaigns 5 times a day and making reactive changes hurts more than it helps.

Here is the 30-minute version for UGC.

Why one routine beats constant checking

That RevenueCat post is written for paid campaigns, and its author is blunt: below $500 a day, make changes twice a week at most, because the budget can't feed more.

The author's week, as published. Monday, 1 to 2 hours: refresh the reports, compare with the previous week, decide what to pause, mine the winners for ideas. Tuesday, 3 to 4 hours: plan and produce a batch of 3 to 5 new assets. Wednesday, 1 hour: check Monday's uploads, pause what didn't spend or perform, upload 1 or 2 from the batch.

Thursday, 20 minutes: refresh, pause bad performers once they have 10,000 or more impressions. Friday, 1 hour: load the remaining assets and decide whether to scale or cut the budget. Plus 15 minutes a day for two questions: is pacing correct (is the budget spending evenly), and is there an anomaly.

That is roughly 6 to 8 hours across the week, plus 15 minutes a day, for paid. UGC compresses the decision part into 30 minutes because production and creator management live elsewhere in your week (Lessons 05, 06, and 09), so the review only reads and decides.

Two speeds, not one. Eric Seufert of Mobile Dev Memo told Sub Club how he handles new creatives: kill "as soon as it's obviously not a winner," which is "oftentimes very quickly, same day, it could be next day, this is getting no delivery, not a winner." No delivery means the platform isn't showing the ad. His summary: "The winners take time to prove out, but the losers are pretty quick to prove out."

So the review kills fast and scales slowly.

Scale / Iterate / Watch / Kill

Every Ad Creative row gets exactly one label each week. Four labels, no fifth.

Kill. For an organic video: engagement rate below 7.5% and zero clicks by the next review. The 7.5% line is Drew Levan of Sideshift's, from a Superwall interview, with engagement rate defined as (likes + comments + saves + shares) ÷ views.

For a boosted creative: no delivery on day one or two (Seufert), or under 50% of its ad group's spend after two days, which a RevenueCat post on creative testing calls a likely loser or false positive. For a Spark Ad: engagement below 5%, Drew Levan's kill line.

Watch. Engagement at or above 7.5% but too few clicks or installs to call, or a video still climbing. Coconote's Zack Hargett described on Sub Club how they read a 20 to 30 million view video: you know exactly when it is going viral, so you measure down-funnel impact on those couple of days.

A Watch row gets one more week. Two at most, then it becomes Iterate or Kill.

Iterate. One stage is broken and the rest work. High views and low clicks means the hook works and the CTA doesn't (Lesson 04). High clicks and low installs points at the link or landing (Lesson 03).

Start Trial without Subscribe is a paywall or onboarding problem, so flag it and move on. The RevenueCat creative testing framework post shows Iterate as a habit: each winner got its own spreadsheet tab of variations, each scored against the original on the metric it was meant to move. Change the hook, expect hook rate (the share of viewers who stay past the opening) to rise.

Scale. Engagement at or above 7.5% with Start Trial and Subscribe present. Organic: boost from the creator's account (Lesson 13).

Already boosted: promote a test creative to the main ad group when its cost per trial (CPT) is at or better than baseline after 10,000 impressions and it is reaching the target audience, and demote a main-group winner when its CPT runs 20% above baseline over the last 2 days. Both rules come from the same RevenueCat post, whose author reports testing 30 or more creatives and finding at least 2 to 3 winners a week.

The testing matrix you carry between weeks

The Tuesday batch should not be random. Keep one table, angle by hook, with the creator and one primary metric in each cell.

One primary metric per cell (Lesson 02). An empty cell is a production order, and iterations of Scale rows fill cells first.

The 30-minute agenda

Same day, same time, every week. Monday matches the RevenueCat routine; any fixed slot works.

1. Minutes 0 to 5: open the saved report. Run the last 7 days and set last week's numbers beside it in your matrix, per Ad Creative.
2. Minutes 5 to 15: label every row. Kill first, because losers prove out quickest. Then Scale, then Iterate. The rest is Watch, with a week counter.
3. Minutes 15 to 22: boost decisions. For each Scale row still organic, apply Lesson 13: wait a day or two if still climbing, then $20 to $100 a day from the creator's account for one week. For boosted rows, apply the promote and demote rules.
4. Minutes 22 to 27: fill the next batch. 3 to 5 assets, the RevenueCat weekly cadence. Iterations of Scale rows first, then empty cells.
5. Minutes 27 to 30: write the sentence and log it. One line, dated, in the same file as the matrix.

If you are boosting, keep the RevenueCat post's 15-minute daily pacing check and its automated rule: "pause this ad group if CPA exceeds $X over the last 3 days" (CPA is cost per action). If you are organic only, skip the daily check: the week is the unit.