Creative Testing Frameworks: How Many Ads You Should Actually Be Running

'How many ads should we be running?' is the most common question we get — and the wrong first question. The right one is: how many distinct concepts are you testing per month, and what's your system for iterating on the winners? Here's the framework we use with brands and creators.

Think in concepts, not ads

A concept is a distinct idea — a hook, an angle, a story. Variations are executions of that concept. Most underperforming accounts run one concept in twelve slightly different outfits. Strong accounts run several genuinely different concepts and let the platform find the winner.

A simple monthly cadence

• Launch 4–6 new concepts per month, each with 2–3 executions (formats, hooks, lengths).
• Give each concept enough budget and time to exit the learning phase before judging it.
• Kill decisively. Double down on winners with fresh iterations rather than nursing losers.
• Keep a backlog. The pipeline should never run dry — creative fatigue is relentless.

Read results at the right altitude

Don't over-optimize on a single metric or a single day. Look at hold rate and hook rate to diagnose creative, and cost-per-result to diagnose the offer and funnel. A weak hook is a creative problem; a strong hook with weak conversion is usually a landing-page or offer problem.

The winner is a starting point, not a finish line

When a concept wins, that's the beginning. Spin it into new hooks, new formats, new lengths. Your best-performing idea should spawn a family of ads, not sit untouched until it fatigues. This is how a testing program becomes a compounding creative pipeline — and it's the core of how we run advertising at Sequence.