Five years ago, media buying was about audiences. Interest stacks, lookalikes, layered exclusions. The buyer who could find the right 200,000 people won.
That game is mostly over. Tracking restrictions took away most of the signal, and the platforms responded by making their own targeting better than anything you can build by hand. Broad targeting now beats most manual setups.
So what is left to control? The ad itself.
The ad chooses the audience
When you run broad, the creative does the targeting. A video that opens with "if you run a plumbing company" will be shown to plumbers, because plumbers are the ones who stop scrolling. The platform learns who responds and finds more of them.
This means creative is no longer a "make it look nice" job. It is the primary lever on who you reach and what it costs.
Stop looking for a winner. Build a system.
The accounts that decay every few weeks are the ones that found one good ad and rode it until it died. The accounts that keep growing are the ones with a production cadence:
- A fixed number of new concepts every month — not "when we get around to it"
- Angles, not just formats — the same offer framed around price, speed, trust, and fear of missing out are four different ads
- A clean testing structure so you know which concept won, not just that something did
- A library — winners get documented, iterated, and rebuilt in new formats
What good looks like
For most accounts spending $10k–$50k a month, that is eight to twelve new ads a month across static, UGC, and short video, with the two or three winners from last month being remixed alongside them.
It sounds like a lot. It is less than most brands spend on the media itself, and it is the difference between an account that compounds and one that needs rescuing every quarter.
- paid-media
- creative