Blog · November 3, 2026 · 5 min read

Brand-name headlines scored lowest in every ad we tested. Fixing them alone didn't sell more.

By Pavan, Founder, FirstReaction · Updated

In short

In five live ads, headlines that were just the brand name scored lowest (34 to 40), while headlines naming a benefit or a specific detail scored 71 to 78. But rewriting only the headline left likelihood to act flat: the ads that sold better changed what the creative showed and proved, and the headline then said it clearly.

Across the five live ads we simulated for What's Wrong With This Ad?, one pattern was hard to miss: the weakest headlines were the ones that only said the brand's name. Rewriting them roughly doubled their scores. Then we tried changing only the headline, and the ad didn't get any better.

Both results are worth knowing before your next headline test.

The headlines, scored

FirstReaction scores the headline, primary text and call to action separately, for each simulated buyer type. Here's the headline score, averaged across buyer types and runs, before and after each rebuild:

AdOriginal headlineScoreRebuilt headlineScore
Sahajan cleansing oil (video)"Sahajan Skincare"34"Melts Makeup. Keeps Moisture."76
Snitch plus-size jeans (carousel)"Snitch"39"Snitch Plus Jeans · Waist 40–50 · Flat 10% Off (App Exclusive)"76
Lola Blankets, Refinery29 feature"Shop Lola Blankets"40"Life-Changing Softness"71
Lola Blankets, holiday offer"Get 35% Off With Code: HOLIDAY35"52"Life-Changing Softness: 35% Off"76
Bella Vita perfume bundle"Buy any 4 Perfume at ₹1599"66(unchanged)74

Three things stand out:

  • Brand-only headlines scored lowest, 34 to 40. A name tells a stranger nothing they can want.
  • A discount alone did a little better (52), and a concrete price-and-quantity offer better still (66).
  • The best headlines said what you get: a benefit ("Life-Changing Softness"), a pair of outcomes ("Melts Makeup. Keeps Moisture."), or the specific details a buyer is checking for ("Waist 40–50").

The twist: the headline alone didn't sell more

For Sahajan, we ran a control: the same video, with only the headline and primary text rewritten. The headline score jumped from 34 to 73. Likelihood to act went from 59 to 56. That's within the normal run-to-run variation, so call it flat, but it certainly didn't rise.

Two more results point the same way:

  • Snitch's first rebuild kept "Snitch" as the headline (it scored even lower, 27) and still raised likelihood to act from 61 to 72, because the carousel cards were fixed.
  • Bella Vita's headline didn't change at all, but its score rose from 66 to 74, because the image and copy around it got better. Headlines are read in context, not alone.

In other words, the headline is the label on the box. A better label helps people understand what's inside, but it can't make them want what's inside. The ads that sold better changed what the creative showed and proved; the headline then said it clearly.

What this means for your headline tests

  1. Fix the creative and the proof first. If the image or video doesn't show the product doing its job, a new headline is polish on the wrong thing.
  2. Then make the headline say what you get. Name the benefit, the outcome, or the one detail buyers check for. Keep the brand name for the logo.
  3. Judge headline tests on likelihood to act or clicks, not on a headline score. A headline that "scores" better can still leave the ad flat.
  4. Test several at once. A bulk test that pairs each headline with the same creative is the quickest way to see which wording adds anything.

Headline patterns that scored well

  • Benefit: "Life-Changing Softness"
  • Two outcomes: "Melts Makeup. Keeps Moisture."
  • The buyer's checklist: "Plus Jeans · Waist 40–50 · Flat 10% Off"
  • Benefit plus offer: "Life-Changing Softness: 35% Off"

And patterns that scored poorly: the brand name on its own, "Shop [brand]", and a discount code with no reason to want the product.

For the other gaps these ads shared, read we simulated 5 live DTC ads: the same 4 mistakes kept showing up. For testing hooks and openings, see the guide to testing Facebook ad hooks.

Sources

  1. Meta Ad Library