How do you pick which ad creative to launch first?

tiktokguru

For anyone running paid ads or working in performance marketing: how do you decide which creative gets the green light first?

Let's say you've got five versions of an ad - different hooks, visuals, copy, thumbnails, CTAs. Before spending a single pound of media budget, what actually determines which one goes live?

In my experience, it's almost always a mix of:

📊 Past campaign data - what drove conversions last quarter
🧠 Gut instinct - that internal feeling that something "just works"
👥 Client preference - often the loudest voice in the room
🛠 Platform best practices - Meta says square video, so square video it is
🔁 Early A/B testing - launch 2-3 variants and let the data sort it out

Honestly, the process is way less scientific than most people think. A lot of the time it's whatever can get approved fast enough. Real testing happens after launch.

What I find painfully common: the creative everyone internally loved absolutely bombing once it hits the feed. Probably happens 40% of the time. Why? Because your team has seen that ad twenty times, knows the strategy, the product, the insider jokes. The customer sees it once while half-scrolling and gives it maybe two seconds. Totally different environments.

So how do you actually make that call? Is it mostly data, instinct, or just whoever shouts loudest in the review meeting?

communityfirst

It's rarely as methodical as we'd like to believe. Most of the time it's a blur of past performance data, creative instinct, and whatever can actually get through the approval chain quickly. The real testing only kicks off once it's live.

I've seen campaigns that the whole team adored internally completely flatline. When you've seen the ad twenty times, know every strategic rationale behind it, and understand the product inside out, you lose all objectivity. The customer sees it once, mid-scroll, and gives it maybe two seconds. Those are two completely different realities.

That disconnect is why I push for rapid iteration over internal popularity contests. Trust the data from the real world, not the conference room. It's humbling every time, but it keeps us honest.

digitalnomad

Oh absolutely, this is such a familiar dynamic. Our process usually looks something like this:

📊 Past data first - dig into which hooks and formats actually drove results before. That's the non-negotiable starting point.
🤔 Gut instinct second - but then everyone brings their own "feelings" to the table, and the internal review becomes a full-blown debate over which visual "lands" best.

And the painful truth? That internal favourite bombs as soon as it goes live way more often than we like to admit - honestly feels like it happens around 40% of the time. You'd think we'd learn, but somehow the cycle repeats 😅

bouncekiller

Honestly? Past data plus a gut check - then launch the top 2-3 and let A/B testing decide which one lives.