During BFCM, you don't have the time to reshoot when a hook underperforms on day two of a five-day sale, and you don't have the luxury of finding out on Black Friday itself that your only asset for a particular persona isn't landing.
<aside> ⭐
The fix is a mindset shift we'd sum up as: don't build ads, build toolkits.
</aside>
Instead of briefing one finished ad, you brief interchangeable pieces - hooks, bodies, and CTAs that can be swapped, reordered, and reassembled without ever going back into production.
Persona research is only valuable if it makes it into the brief this way.
Build variation in from the start, so every piece of creative you produce is engineered to reach a different persona, generate a different signal, and give the algorithm a genuinely new data point (not just a resize of the same idea).
Remember, when you find something that strikes a chord with a specific persona, that's the valuable part - the gold. Rebuild the vehicle carrying that insight across every format and creator identity that persona might respond to.
One strong insight —> five different executions, is worth more during BFCM than five different insights shot once each.
Structure every brief into three interchangeable sections, each with 3 - 5 variations:
Every hook should combine three elements:
Pattern interrupt: something that breaks the feed. Avoid familiar, polished, ad-like visuals; the more it looks like content, the longer it holds attention.
Angle: hit a pain point or benefit specific to this micro-persona, not a generic one that could apply to anyone.
Curiosity loop: create enough tension that the viewer has to keep watching to resolve it.
Pattern interrupt + angle + curiosity loop = a winning hook.
Build 3–5 hook variations per persona, all pointing at the same underlying insight, so you've got options ready the moment one starts to fatigue mid-sale.