The brand made custom metal wall art, family names cut into steel. It had been stuck around ten thousand a month for a long stretch, barely breaking even, while a competitor with a charismatic founder ate the feed.
The team's read was that they needed a face. They had been copying the competitor's founder-personality angle for months.
Nobody had asked who was buying.
So I exported every Facebook campaign to CSV and built cohorts in Google Sheets. Creative, age, device, headline. Boring work, a few afternoons of it.
One pattern came out immediately. Women, forty-five and up, on iPhones, buying as gifts.
That is a completely different customer than the one the ads were written for. The ads were art-direction-forward, addressed to design-conscious millennials who clicked and never bought. The actual buyer was not shopping for design. She was buying a thing that would mean something on a wall, for a person she loved.
We rebuilt the offer around her. Family is Forever. Make Your House a Home. Same product, same factory, same photographs in some cases. New promise.
Then we split the campaigns by age and gender so budgets and bids could follow the buyer instead of chasing an average that described nobody.
Ninety days later the brand was at a hundred and fifty thousand a month. Within about seven months that line had crossed a million, and the company eventually grew into an eight-figure operation with its own team in house.
Segmentation beats swagger.
The lesson I keep taking from this one is how unglamorous the winning move was. No clever hook, no new creative direction, no founder on camera. A CSV export, a pivot table, and the willingness to find out that the customer we had been writing to did not exist.
Most brands stuck at a plateau are not short on ideas. They are writing to an imagined buyer and have never checked the imagined one against the real one.