What if I don't have a winning video yet?

Start with a variety of new ideas and test them with small, equal budgets. The first video that hits your goal becomes the starting point for the 60/40 rule.

Does iteration bore the audience?

Smart iteration changes the face, the format, or the segment, so it looks like a new ad to the audience even though the idea is proven. What bores the audience is showing the same video for a long time.

How many versions should I make of a winning video?

At least three, one from each axis. Whichever version succeeds becomes the start of a new round of iteration.

Iterate on your winning video: the 60/40 rule and three ways to iterate

Most accounts chase a new idea every week, and most new ideas don't work. The smarter move is to build on the video that has already proven itself. This guide shows you how to iterate on it in ways that extend its life and uncover even stronger versions.

Abdulrahman Yaqut · Last updated: September 26, 2026

Make 60% of your weekly production new versions of proven videos and 40% new ideas. Iterate on the winner along three axes: the same script with a different creator, the same message in a different format, or the same idea for another customer segment. And set a spend limit in advance at which you stop any version that isn't hitting your goal.

Why is iterating smarter than inventing?

A completely new idea is an unknown: you don't know whether the hook grabs attention or whether the message persuades. A winning video, on the other hand, has proven that its idea works with your audience. Every new version starts on tested ground, so its odds of success are much higher than an untested idea.

Performance-advertising expert Dara Denney calls this approach the "lazy strategist": creativity isn't about inventing something new every week, but about iterating on what works, intelligently.

The 60/40 rule

Split your weekly production like this:

60% iteration: new versions of videos that have proven themselves in your account.
40% new: ideas and angles you haven't tried before, to discover the next winner.

Most accounts do the opposite: most production goes to new ideas and little to iteration. The result is that the winning video wears out quickly with no ready replacement, and spend stops growing.

Three ways to iterate on a winning video

Same script, different creator: the easiest iteration and usually the most successful. Give your best video's script to three creators from the same category but with different styles. One of them often outperforms the original. Find them easily with AI search.
Same message, different format: was the video a creator talking to camera? Try it as product shots with a voiceover, a before-and-after story, or an image carousel. Same message, different template.
Same idea, another segment: did the video work with young people? Make a version that speaks to parents. Did it work with women? Try a version for men if the product suits them. Same problem, in a new segment's language. To reach expats in their own dialects, see expat communities.

Example: a winning perfume video, with a creator talking about how the scent lasts through the workday. Iterations: the same script with two other creators, a version with close-up shots of the perfume and a voiceover, and a version aimed at people shopping for a gift instead of for themselves.

The fastest iteration: from video hook to image ad

This iteration needs no new filming. Many people scroll with the sound off, and an image ad is cheaper, faster, and clearer in its message:

Sort your videos by spend and take the top three.
Write down what is said and shown on screen in the first three seconds of each video.
Boil the essence of the hook down to 3 or 4 sharp words with a clear promise.
Put it in large type on a simple background, in your brand color or with a product photo, with a clear call to action and a small logo.

When to stop a version that isn't working

Not every version will work, and that's normal. What matters is setting a spend limit before launch for judging each version, for example two or three times your target cost per acquisition. If a version reaches that limit without getting close to your goal, stop it, with no "last chance."

This limit protects your budget, teaches you quickly what doesn't work, and directs spend to the versions that succeed. To read the results correctly, see How to test ad videos.

A typical production week

Start of the week: pull the top 3 videos by spend over the last 14 days, plan nine versions of them (3 axes per video), and write the briefs.
Midweek: film the versions with creators, and prepare 3 to 5 image ads from the winning hooks.
Before the end of the week: 3 or 4 completely new ideas, which is the 40% share.
End of the week: launch all versions in one test, review them the following week, and stop whatever exceeded its limit.

In UGCeer, performance analytics shows you the winning video and why it works, and you can order its next version from the same creator or a new one in a click.