Thumb-Stop Ratio is the share of users who stopped scrolling on an ad long enough to register a 3-second view, calculated as 3-second views divided by reach (not impressions).
Thumb-stop ratio measures what fraction of unique users actually paused their scroll long enough to log a 3-second view on your ad. The denominator is reach — not impressions — which is the critical distinction separating it from hook rate.
The mechanism is straightforward: a platform counts every unique user who was served the ad (reach), then counts how many of them crossed the 3-second threshold. Divide the second by the first. The result tells you how compelling your leading frame is on first contact. Because each user is counted once, the metric doesn't inflate with repeat exposures the way impression-based rates do. A high-frequency campaign that keeps serving the same creative to the same people can produce a flattering hook rate while its thumb-stop ratio quietly declines — a signal worth catching before creative fatigue sets in.
For prospecting campaigns on TikTok and Meta Reels, this metric is one of the cleaner leading indicators available. It answers a single question: does this frame earn the next two seconds? In 2025–2026, with Andromeda routing impressions across placements automatically and Advantage+ Creative adapting assets on the fly, the first frame carries more weight than ever. The ad creative that wins the thumb-stop is the one that gets the downstream optimization budget.
Context matters for benchmarks. Creator-led UGC formats on TikTok typically hit 35–45% in cold audiences; brand-produced spots rarely clear 25%. I treat anything below 18% as a rebuild trigger on the opening frame, regardless of other metrics. The data behind these patterns is covered in depth in ai-for-tiktok-ads-2026 and best-ai-ugc-video-tools-2026.
The practitioner principle: track thumb-stop alongside conversion rate — a 40% stop rate attached to 1% purchase conversion is a different problem than a 22% stop rate with 4%.
Hook rate uses impressions; thumb-stop uses reach. The reach-based denominator filters out repeat exposures, giving a cleaner read on first-impression strength. At scale, the two metrics diverge — thumb-stop is the truer signal. I find it particularly useful in retargeting audits: when frequency climbs above 4, hook rate holds steady while thumb-stop often drops, exposing the gap between repeat-viewer behavior and genuine new-attention capture.
