Stop Making New Ads. Start Mining Your Winners.

When performance dips, the reflex is to brief something new. New scripts, new shoot, new everything.

Sometimes that’s right. But most of the time, the cheapest and fastest win in the account is the ad that’s already working. A winner isn’t one ad. It’s a proven message with a dozen variations you haven’t tested yet.

Our monthly split: More, Better, New

Every month’s creative on an Aura account is split roughly like this:

• About 50% “More”: iterations of current winners.

• About 25% “Better”: stronger rebuilds of ads that almost worked.

• About 25% “New”: concepts that have never been tested.

New concepts are how you find the next big winner. Iterations are how you pay for the search.

The one-variable rule

Every iteration changes exactly one thing:

• The hook: new first three seconds, same body.

• The body: same hook, different proof or demo.

• The visual: same script, different creator, setting, or look.

• The format: same message, new wrapper (talking head becomes a static, a static becomes a carousel).

Change two things and a result tells you nothing. Change one and every test teaches you something you can use next month.

Where to find iterations hiding in a winner

• Unused openings. Watch the full video and mark every moment that could be a first frame. There are usually several.

• Shorter cuts. A 45 second winner often has a 15 second version that does even better in Reels and Stories.

• Different creators. Same script, new face. This is also how you get real visual variety, which Meta rewards.

• Static versions. The strongest line in the video becomes the headline of a static.

• New audiences. The same proof, reworded for a different buyer or use case.

Near-misses are worth a second look

An ad with a strong hook but weak conversions has a body problem. An ad that converts well but nobody stops on has a hook problem. Both are closer to winning than a blank page. Diagnose which half failed, fix that half, and relaunch.

Do this this week

1. Sort your last 30 days of ads into three groups: winners, near-misses, and dead.

2. Pick your top two winners and write four one-variable iterations of each.

3. Pick one near-miss and decide whether it failed on the hook or the body.

4. Save everything else for “New.”

Want this run on your account?

The 30-Day Profit Pilot puts this system on your Meta account for one month: account rebuild, a fresh batch of tested creative, and a weekly scorecard tied to your CPA. If it doesn’t make sense for where you are, we’ll tell you on the call.