TikTok Ads: what actually works, and what transfers from Meta
The most common way to waste money on TikTok is to run your Meta creative on it.
The offer transfers. The audience logic transfers. The format, pacing and tone do not — and on TikTok, format is most of the performance.
The platform's actual rule
TikTok's feed rewards content that looks like content. The moment something reads as an advertisement, the swipe happens before your message does.
This isn't a stylistic preference. It's mechanical. You are competing for attention against organic video made by people whose entire job is holding attention, and the audience has trained itself to identify and skip ads within about half a second.
What this means practically
Production value works against you. A colour-graded studio shot signals "ad". A phone-shot clip in a real room signals "person". The second performs better, consistently, and it's cheaper to make.
The first second is the entire game. Not the first three seconds — the first one. Movement, a face, a caption that creates a question, or an immediate visual payoff.
Sound matters more than on any other platform. Voice, trending audio, natural ambient noise. Silent-first design that works on Meta fails here.
Native captions and text. Use TikTok's own visual language rather than importing brand templates.
Creative structures that work
The demonstration. Show the thing doing the thing. No setup, no branding, straight to the product working.
The problem-reaction. Open on the frustration, not the solution. The viewer recognises themselves in the first second.
The talking head with a claim. Someone saying something specific and slightly contrarian directly to camera.
The before/after. Only works if the "before" is honest enough to be believable.
The list. "Three things I'd change about..." — structurally satisfying, easy to hold attention through.
What transfers from Meta
- The offer. If the offer doesn't work on Meta it won't work here.
- The core hook logic. What makes someone care is the same; how you deliver it isn't.
- Audience economics. Your CPA target, margin constraints and LTV maths don't change.
- Testing discipline. Concept-level testing works the same way — see the Meta creative testing framework.
What doesn't
- Creative assets, directly. Recut at minimum. Usually reshoot.
- Creative lifespan. TikTok creative fatigues faster. Plan for shorter cycles.
- Comment culture. Comments are part of the ad on TikTok. They get read, they influence perception, and they need moderating.
Spark Ads
If you have organic content that performed, running it as a Spark Ad — promoting the actual post rather than a fresh upload — usually outperforms. It carries existing engagement, it looks native because it is native, and the social proof is real rather than implied.
This is the single highest-leverage TikTok tactic for most businesses, and it's underused.
Who should skip TikTok
Be honest about fit:
- Long, complex B2B sales cycles. Possible, rarely efficient.
- Businesses with no capacity to produce video regularly. TikTok consumes creative faster than any other channel. Without a pipeline you'll spend two weeks setting it up and have nothing to feed it by week three.
- Anyone whose tracking can't distinguish channels. You'll never know whether it worked.
The realistic starting point
Four to six native-style videos. Broad targeting. Small daily budget. Two-week window. Judge on hook rate first, then CPA.
If nothing holds attention past the first second, the problem isn't targeting or budget. It's the creative, and no amount of optimisation fixes that.