What Makes an Excellent Ad

An excellent ad isn’t the one that wins awards. It’s the one that scales profitably for months. We’ve seen the pattern across hundreds of ads run for clients. It comes down to 4 components.

1. Hook that stops the scroll

The first second decides everything. If you lose 50% of viewers by second 2, the budget is wasted. The hook isn’t just text on screen. It’s the promise that answers the question “why should I stop?”.

At Old Spice we tested 11 hooks on the same body. The difference between best and worst was 7x on cost per completion. The content was identical. Only the hook was different.

2. Clear, personalized problem

After the hook, in the first 5-8 seconds, the viewer has to understand: “ok, this is about me”. If they don’t recognize themselves there, they’re gone.

Classic mistake: brands describe the product. Wrong. You describe the problem. At Mondly, the best ads didn’t say “learn languages”. They said “you forget the words you learned 3 days ago”. Specificity makes the difference.

3. Solution demonstrated, not declared

You don’t say the product solves it. You show it. Demo, before/after, real result. The 2026 audience doesn’t believe generic claims.

At Gillette we abandoned everything that was generic narrative voice-over. We switched to clips where we show concretely the texture, the motion, the user reaction. CTR went from 1.1% to 3.7% in 8 weeks.

4. Specific and urgent CTA

“Learn more” isn’t a CTA. “Visit the site” isn’t either. A good CTA has a concrete verb + specific benefit + timing. “Start the free lesson today”. “Download the template in 30 seconds”.

At Peugeot we tested 6 CTAs on the same ad. The one with a concrete number (“Book a test drive in 2 minutes”) had double the CTR of the generic one (“Find out more”).

What an excellent ad does NOT do

It doesn’t have a voice-over that sounds like an ad. It doesn’t have generic stock music. It doesn’t have a big logo in the first 3 seconds. It doesn’t have “values” as the central message. It doesn’t have 30 seconds of intro for an 8-second message.

At Citroen we removed all these elements from a campaign. Result: ROAS from 1.8 to 4.3. Same budget, same audiences.

Testing: how you find excellence

You don’t create it, you discover it. Formula: 10 variants, run with small budget (50-100 euros each), 3-5 will be obviously better. You scale those 2.

Brands that try to make “the perfect ad” on the first try spend more and get less. Iteration beats intuition.

A simple review framework

Before publishing an ad, answer 4 questions: Do the first 2 seconds stop the scroll? Do the first 8 seconds clearly say what it’s about? Am I showing or just telling? Does the CTA have verb + benefit?

If any answer is “no”, the ad isn’t ready. Check on TikTok Ad Librarythe competitor ads that have been running for 30+ days. There you’ll see how they apply the framework.

Excellence in advertising isn’t art. It’s discipline applied across many iterations. Brands that understand this make more money with smaller budgets.