← All tools

Free tool · no signup

Which hook stops the scroll?

Paste two hooks or titles. We score both on 8 factors and crown a winner — live as you type.

Hook A/B Tester

42/100

20/100

Platform
How we calculate this

Each hook is scored out of 100 across 8 factors:

  • Length sweet spot (6–12 words): +20 — drops to +12 at 13–16 words, +8 under 4, +6 beyond that
  • Curiosity words: +13 each, capped at +25 — why, how, stop, mistake, actually, nobody…
  • Specific number: +15 — "3 mistakes" beats "some mistakes"
  • Question mark: +10 — a question opens a loop viewers want closed
  • Front-loaded verb: +15 — start with stop, watch, try, learn…
  • ALL-CAPS penalty: −12 when over 60% of letters are capitalized
  • Emoji count: +10 for 1–2, +5 for none, +2 for a spam wall
  • Platform fit: +5 — TikTok likes 5+ words, YouTube likes a number or question, Instagram likes emoji restraint

This is a heuristic estimate — it scores the words, not your delivery. Real watch-time data always wins the argument.

Head-to-head on TikTok

The 1.5-second decision

Every scroll starts the same way. A viewer gives your video about 1.5 seconds — one glance at the first frame, one half-read of the first line — and then their thumb decides. Your hook isn't the intro. It's the whole audition.

That's why creators searching for a TikTok hook generator or a YouTube title tester are really asking the same question: which words make someone stop? The answer is rarely "cleverer words." It's structure. Hooks that win are specific, open a loop, and promise a payoff the video actually delivers.

Open loops vs clickbait — where the line is

An open loop teases information the viewer can only get by watching: "the mistake that killed my reach" works because the viewer wants it closed. Clickbait runs the same play with no payoff — the loop never closes, watch time collapses, and the algorithm learns not to trust you. The line is simple: a hook is a promise. Keep it and it's a hook. Break it and it's clickbait. Your retention graph knows the difference within seconds.

5 hook formulas that work across niches

  • The numbered mistake: "Stop making these 3 editing mistakes." A number plus the fear of missing the fix. Concrete, scannable, works from cooking to coding.
  • The contrarian why: "Why nobody should start a podcast in 2026." Pick a fight with common advice and the comments do your distribution for you.
  • The secret / reveal: "The setting pros use that nobody talks about." Curiosity words like "secret," "actually," and "nobody" pull hard — as long as the reveal is real.
  • The stop-doing-X: "Stop posting at 9am. Do this instead." A front-loaded verb, instant tension, and an implied better way in under ten words.
  • The how-I-got-Y: "How I got 10,000 followers with 12 videos." A specific result plus an implied roadmap. The more concrete the number, the harder it pulls.

Test, don't guess

Scoring hooks here is step one. Step two is real data: post both versions — or run YouTube's built-in title test — and compare 3-second hold rate or click-through rate. One variable at a time, enough impressions each. Keep a swipe file of what wins in your niche; hooks compound faster than almost any other skill. Then post at the right time so the winner gets a fair first hour.

Scores are a heuristic estimate of text strength, not a guarantee. Delivery, visuals, and the first three seconds of footage matter just as much.
What makes a good TikTok hook?

A good TikTok hook opens a loop in the first 1.5 seconds — a specific promise, a curiosity gap, or a pattern interrupt. Numbers, curiosity words like 'why' and 'stop', and a front-loaded verb all help. Keep it 6–12 words, then deliver the payoff fast: watch time decides reach.

How long should a hook be?

Six to twelve words is the sweet spot for a spoken or on-screen hook — long enough for a specific promise, short enough to land before the thumb moves. On YouTube, keep titles under about 60 characters so they don't truncate in search and suggested.

Are questions good hooks?

Yes — when they open a loop the viewer genuinely wants closed, like 'why does nobody talk about this?'. Rhetorical questions with obvious answers hold nobody. Pair the question with a specific payoff and answer it early in the video.

Should I use ALL CAPS in titles?

Sparingly. One capped word for emphasis can work on YouTube, but a full ALL-CAPS title reads as shouting and tends to underperform. This tool docks hooks that are more than 60% capital letters.

How do I test hooks in practice?

Post the same content with different hooks across a week — or run YouTube's built-in title test — and compare 3-second hold rate on TikTok and Reels or click-through rate on YouTube. Change one variable at a time and give each version enough impressions before you call a winner.