If we listed the things that determine how an ad creative performs, what would be at the top?
Not the quality of the product photography. Not the correct use of brand colours. Certainly not how cleverly the copy is written.
At the top are the first two seconds.
Because if an ad cannot get past the first two seconds, nobody watches the rest of it. How good the rest is does not matter — nobody sees it.
The scroll speed problem
The average Instagram user passes three or four pieces of content a second. On TikTok it is faster still.
Which means: by the moment your ad appears on screen, the user’s thumb is already moving. Your job is to stop that movement.
Anything that does not stop it — however beautiful — counts for nothing in performance terms.
This is why “beautiful creative” and “creative that works” are different things. Award-winning ad films are usually bad performance ads, because telling a story takes time. You do not have that time.
An ad that looks like an ad loses
This is the most basic rule.
The moment a user recognises a piece of content as an ad, they go on the defensive. It is a learned reflex, developed over years of avoiding advertising.
Signs of looking like an ad:
- A product photo isolated on a white background
- Logo in the corner, tagline underneath
- Big “50% OFF” text
- Stock photo smiles
- Editing that is too clean, colours that are too bright
Signs of not looking like an ad:
- Footage shot on a phone
- A messy background (a real home, a real desk)
- A screen recording
- A person talking
- Text appearing on screen as if handwritten
The second group looks “worse”. And it usually works better.
The eight families of winning hooks
In the creative we produce we sort openings into eight families. When we start production for a new brand we make at least one example from each — because there is no way of knowing in advance which will land.
1. Confession. “If only I had known this 6 months ago…” — A personal mistake or discovery. It creates curiosity.
2. Comparison. “The left one is 300₺, the right one 1,200₺.” — Two things side by side. The eye automatically looks for the difference.
3. Mistake. “If you are doing this, stop.” — A direct warning. The viewer wants to check themselves.
4. Transformation. “Day 1 / Day 30” — Before and after. One of the oldest patterns, and still one of the strongest.
5. Price shock. “Guess how much this costs?” — It makes people guess. They stay to find out the answer.
6. Process. “Watch how it is made.” — Production, preparation, behind the scenes. People love watching a process.
7. Reaction. “Look at these comments…” — Someone else’s reaction. Social proof at its fastest.
8. List. “3 things, 30 seconds.” — It says up front how long this will take. That earns a promise to stay.
I would not suggest memorising this list. What I would suggest is asking, as you produce your next batch of creative, “which family is this one from?” If they all come from the same family you are not testing — you are making eight versions of the same thing.
How big is the creative difference in numbers?
The performance gap between two creatives in the same campaign, shown to the same audience, is typically around threefold.
That figure is worth dwelling on. It means that even if you perfect your campaign structure, your targeting and your bid strategy, you cannot close the gap between bad creative and good creative.
The average change we see in the accounts we manage after a creative refresh:
- Click-through rate: +180%
- Cost per thousand impressions: -42%
- Customer acquisition cost: -35%
The CPM falling may look odd at first — what does creative have to do with the cost of impressions? This much: platforms reward content people engage with. If people stop, watch, like and comment, the platform distributes that ad more cheaply. Good creative makes your advertising cheaper.
What to do once you have found the winner
The most common mistake: find the winning creative, pile budget on it, and wait.
A winning creative tires too. Six to eight weeks on average (two to three on TikTok). Then CTR starts to fall.
The right move, once you have a winner, is to build variations around it:
- Same content, different opening line
- Same content, different first frame
- Same content, different music
- Same content, different closing call to action
- Same message, different person on camera
That way what wins is not a creative but a concept. A concept lives far longer than a single video.
The question of production volume
Everything above depends on one thing: are you producing enough creative?
A rough calculation. Around 10-15% of the creatives you produce fall into the “works” category. The share that is genuinely good and scalable is 3-5%.
So if you produce four creatives a month, it will take five to eight months on average before you stumble on a scalable one. By then your existing creatives will have tired anyway.
Produce 16 a month and you get the same chance every month.
That is the mathematical answer to why creative production is a continuous process, not a “project”.
Summary
- If a creative cannot get past the first 2 seconds, the rest of it does not matter.
- An ad that looks like an ad loses.
- If your openings always come from the same family, you are not testing.
- The creative gap is too large to close with campaign optimisation.
- When you find a winner, do not stop — build variations around it.
- Produce little and your odds are small. This is a volume game.
If you are wondering why your own ads are not stopping anyone, we review your live creative for free — we send you, in writing, what is not working and why.