Guide
On TikTok, the algorithm decides your reach — and the algorithm follows attention. If viewers swipe past your first three seconds, your CPM rises and your delivery dies. Test the attention before you buy the reach.
TikTok is not a captive-audience platform. Viewers are mid-scroll, thumb already moving, and your ad has roughly 1–3 seconds to earn the next 10. TikTok's own delivery system amplifies this: ads that hold viewers get cheaper distribution; ads that get swiped get throttled. A weak hook doesn't just lose one viewer — it raises the cost of every impression that follows.
PreTestAds scores your ad against a benchmark of 76 top-performing ads curated from the TikTok Creative Center — the creative TikTok itself showcases as exemplary. Your score is a percentile: 0.72 means your ad's predicted engagement profile beats 72% of those proven top-performers.
That makes the score directly relevant to the platform you're spending on — not a generic attention heatmap, but a comparison against ads validated by TikTok's ecosystem. The same short-form instincts carry to other swipe feeds — see Instagram Reels ad testing and YouTube Shorts ads.
Write three hook variations of the same ad (see how to write a TikTok hook) and cut each one. Score all three — minutes each, no ad spend. Take the winner, fix its attention drop, and launch it as a Spark Ad with your test budget. You enter the auction with creative that has already survived a screen, instead of paying TikTok to eliminate your weak variants. See how predicted engagement relates to ad performance and our full methodology.