Creative is the new targeting: how to run Meta Ads now

A few years ago, the craft of Meta Ads lived in the targeting: stacked interests, lookalike percentages, carefully excluded audiences. Today most of that work happens inside Meta’s own delivery system. Broad targeting and Advantage+ campaigns often match or beat hand-built audiences — because the algorithm sees far more signal than any media buyer can.

That doesn’t make the job easier. It moves it. If the platform decides who sees your ad, your creative decides who stops for it. The ad itself has become the targeting.

Why creative became the targeting

When you give Meta a broad audience, it uses your ad to learn who responds. A video that opens with a founder talking about a payroll headache will find different people than a video that opens with a satisfying product demo — even inside the exact same campaign settings.

Meta has been open about the fact that its newer ad retrieval systems work best with a wider variety of creative to choose from. In practice, every distinct concept is a new way into a new pocket of your market. One great ad can only find one kind of person. Ten genuinely different ads can find ten.

Diversity means concepts, not variations

Most teams hear “more creative” and produce more versions: the same video with a new colour, the same static with a different headline. Those are useful later. But early on, they mostly compete for the same audience.

Real creative diversity comes from changing the things that change who the ad speaks to:

• The persona — a busy founder, a first-time buyer, a team lead, a parent.
• The angle — the pain you remove, the gain you unlock, the proof that it works, the objection you crush.
• The format — talking head, UGC-style review, product demo, carousel, meme, motion graphic.
• The hook — the first two to three seconds that earn the rest.

A useful rule: if two ads would make sense to the same person for the same reason, they’re variations. If they’d land with different people, or for different reasons, they’re concepts.

A simple testing structure

You don’t need a complicated account to test well. A structure we like:

1. One scaling campaign with your proven winners, broad targeting and the budget that pays the bills.
2. One testing campaign where new concepts get a fair, controlled read — enough budget per concept to reach a decision, not so much that one test eats the month.
3. Graduation rules written down in advance: what cost per result, after how much spend, earns a concept a place in the scaling campaign.

Keep the number of campaigns low. Every extra split thins the data the algorithm needs to optimise.

The metrics that tell you why

Cost per result tells you whether an ad works. To know why, look one level down:

• Hook rate — the share of impressions that become three-second views. Weak hook rate means the first frame isn’t earning attention.
• Hold rate — the share of those viewers who keep watching. Strong hook, weak hold usually means the promise of the opening isn’t paid off.
• Click-through rate — whether the ad creates enough desire to act.
• Conversion rate on the landing page — whether the page keeps the promise the ad made.

These turn a gut feeling (“this ad feels tired”) into a brief (“the hook still works; the middle loses people — rebuild the demo section”).

Feed the algorithm clean signal

Creative tells Meta who to find. Your conversion data tells it whether it found the right people. If that data is patchy, the algorithm optimises toward the wrong customers — no matter how good the ads are.

At minimum:

• Run the Conversions API alongside the pixel, with proper deduplication.
• Send the events that predict real value — purchases, qualified leads, trials, subscriptions — not just page views.
• Pass conversion values where you can, so the system learns which customers are worth more.
• Watch your Event Match Quality score and fix what drags it down.

We cover the full setup in our guide to server-side tracking and Conversions API.

A weekly rhythm that compounds

The teams that win on Meta aren’t the ones with a single brilliant ad. They’re the ones with a rhythm:

• Monday: read last week’s tests — concept, hook rate, hold rate, cost per result.
• Tuesday: write briefs for new concepts and iterations on winners.
• Wednesday–Thursday: produce.
• Friday: launch into the testing campaign; graduate last week’s winners.

Run that loop every week and your account gets smarter every week. Skip it and fatigue catches up — usually right when you’ve increased the budget.

Common mistakes to avoid

• Testing too many things at once with too little budget, so nothing reaches a conclusion.
• Judging creative after a day. Delivery takes time to settle; set a minimum spend before you decide.
• Killing ads that work because you’re bored of them. Your audience sees them far less than you do.
• Treating UGC as a style instead of a strategy. A creator video still needs an angle, a hook and proof.
• Letting the landing page break the promise. The ad’s job ends at the click; the page has to finish it.

The takeaway

On Meta, budget gets you into the auction. Creative decides what you get out of it. Build a system that produces genuinely different concepts every week, read them with the right metrics, feed the platform clean signal — and let the algorithm do the targeting it’s now better at than any of us.

If you want an outside view of where your Meta account stands, our free growth audit looks at creative, structure and tracking in one go.