How many UGC ads should your app test on Meta?

Most app founders who try UGC for the first time order one video, run it for a week, and decide whether "UGC works" based on that single ad. That test tells you almost nothing. One video is one hook, one creator, one angle and one edit. If it flops, you have no idea which of those four things killed it.

Short answer: test around 4 UGC ads on Meta to start, split across 2 angles with 2 hooks each, then add new creatives every month. Here is why, and how to read the results.

Why one ad is never a real test

A UGC ad has a few moving parts, and each one can make or break the result:

• The angle: the reason someone should care. "Save time", "stop feeling guilty about X" and "the app my friend showed me" are three different angles for the same product.
• The hook: the first two or three seconds. Same angle, different opening line, very different thumb-stop rate.
• The creator: who is talking. Your audience will trust some faces and voices more than others, and you can't predict which in advance.
• The format: talking head, screen recording with voiceover, a "day in my life" story, a reaction.

When you test a single ad, all four variables are locked together. A win or a loss can't be traced back to anything you can reuse.

A simple starting structure

For an early-stage app with a modest test budget, a good first round looks like this:

1. Pick 2 angles that you genuinely believe in. Pull them from your reviews, support emails and onboarding answers, not from a brainstorm.
2. Write 2 hooks per angle. Keep the body of the script the same so the hook is the only thing that changes. If you are writing them yourself, here is how to write a UGC ad script for an app, with a full example.
3. Use 2 different creators if you can, so one face doesn't decide the whole test.

That gives you 4 core videos (it's also roughly how a UGC production pack is structured), and with hook swaps you can cut 6 to 8 ad variations from them. It's enough to see which angle pulls and which opening lines stop the scroll, without spreading your budget so thin that nothing gets enough impressions to judge.

Test angles first, then hooks, then creators. Each round should answer one question.

How to read the results

Don't judge the round on cost per install alone in the first few days. Look at the signals that tell you which part is working:

• Hook rate (3-second views divided by impressions) tells you if the opening works.
• Hold rate (people still watching later in the video) tells you if the story keeps them.
• Click-through rate tells you if the promise is strong enough to act on.
• Cost per install or per trial is the final verdict, once each ad has had enough spend.

If an angle has a good hook rate but nobody clicks, the opening is fine and the promise is weak. If nobody gets past second three, rewrite the hook before you throw out the angle.

When to make more

Even a winning ad wears out. On Meta, the same audience keeps seeing your best creative, frequency climbs, and performance slowly drops. That's creative fatigue, and it hits small apps faster because their audiences are smaller. For the full weekly check and refresh order, see how to spot and fix Meta ads creative fatigue.

Signs it's time for a refresh:

• Frequency keeps rising while click-through rate slides down.
• Cost per install creeps up week after week with no change in targeting.
• Your "winner" is the only ad getting spend and there's nothing ready to replace it.

The fix isn't one new video. It's a steady flow: take the angle that won, write new hooks for it, try it with a new creator, and add one fresh angle each round so you keep learning.

The short answer

One UGC ad is a coin flip. Around 4 videos split across 2 angles, with a couple of hook variations each, is the smallest test that actually teaches you something. After that, plan to add new creatives every month instead of waiting for your best ad to die.

FAQ

How many UGC ads should I test on Meta at first?

Around 4 videos split across 2 angles, with 2 hooks per angle. One single video can't show you what made it win or lose.

How long should I run a UGC ad test before judging it?

Let each ad get enough spend and impressions to compare them fairly. In the first days, judge hooks with hook rate and hold rate, and keep cost per install or per trial as the final verdict once spend has built up.

How often should an app refresh its UGC ads?

Plan new creatives every month. Refresh sooner when frequency keeps rising, click-through rate drops and cost per install creeps up with no change in targeting.