Which Headline Actually Wins? A Walkthrough
There's an old line about advertising that's been repeated for well over a century:
half the money spent on it is wasted, and the trouble is nobody knows which half.
It's usually credited to John Wanamaker, the department store pioneer — though a fair amount of digging by advertising historians has never turned up a record of him actually saying it. It's been pinned on him, on the soap magnate William Lever, and David Ogilvy hedged his bets by crediting both men in his own writing. Nobody really knows who said it first.
What's true regardless is the problem itself, and it hasn't gone away for small businesses.
A few weeks back on LinkedIn, I posted about a way to actually answer that question without a marketing degree or an ad platform. That post got more engagement than most of what I write, so it's overdue for a proper write-up — here's the full version, with a real example.
The method
Pengi generates marketing copy in five distinct styles for the same promotion — Customer Desire, Problem → Solution, Question, How-To, and Benefit. Same offer, same details, five different ways of framing the headline.
Instead of picking one and hoping it lands, you can generate two or three of them, turn each into its own poster, and actually find out which one works.
Step 1: Generate two or three headline styles for the same promotion. Keep everything else identical — the offer, the discount, the dates, the design. The only thing that changes between versions is the headline framing.
Step 2: Give each version its own code. Not a complicated tracking system — just a short code printed on the poster or mentioned in the flyer, redeemable for the discount already on offer.
Step 3: Run all versions at the same time, not one after another. If one version runs this week and another runs next week, you can't tell whether the difference came from the headline or just from the week — maybe it rained, maybe a competitor ran a sale. Put different versions in different spots, or hand them out in parallel, so the only real difference between them is the headline.
Step 4: Count the codes when the promotion ends. Whichever code came back most is your answer.
A real example
We tested this with Radiance Beauty Studio, a salon, promoting a 10% off consultation offer. Three headline styles, same offer, same discount, same visual layout:
- How-To style: "How to Finally Get the Exact Look You Want — Every Single Time"
- Question style: "Why Do You Keep Leaving the Salon Disappointed?"
- Problem-led style: "Tired of Leaving Disappointed?"
Same 10% off code structure on each poster, each version handed out or posted separately, each with its own code. At the end of the promotion window, whichever code customers actually mentioned or redeemed most told the business which framing genuinely got people through the door — not which one the owner personally liked best, or which one "sounded right" in a meeting.
Worth being honest about
This isn't a scientific A/B test. There's no statistical significance calculation, no controlled sample size, no way to isolate every variable perfectly. A determined marketer with an ad platform and a real budget can do better. But for a small business owner who has never had any way to answer "does this headline actually work," a handful of codes and an honest count at the end of the promotion is more real signal than most ever get — and it costs nothing beyond the time it takes to generate a second version of the poster.
That second version used to be the actual barrier. Making two or three variants of the same poster meant two or three rounds of design work and print runs, which is a lot to ask for a single promotion on a small budget. That's the specific gap Pengi closes — generating a second or third headline style for the same material takes minutes, not a design cycle. Once you actually know which version wins, you're not guessing anymore. You know, and you can put everything into the version that works.