Hook Testing: Write 10 Openers and Pick 2 Before You Publish

The best hook is rarely the first one you write or the cleverest one you write. It's the one that tells the right person exactly what they'll get and why they should believe it. So don't pick hooks by feel. Write 10 openers, score each one from 0 to 2 on five checks (specificity, tension, promise, audience match, proof), and publish with the top 2. It takes about 15 minutes and works on any platform.

This article is about choosing hooks, not writing them from scratch. If you need raw material, our 12 scroll-stopping hook templates and the hook generator guide give you plenty. This is the filter you run those drafts through.

Why "clever" is the wrong test#

When creators pick a hook by instinct, they usually pick the one that sounds best to them: the wittiest line, the one with a pun, the one that feels most like a headline. But the person scrolling isn't grading your wit. In about a second they're deciding whether this post is for them and whether it's worth their time.

Clever hooks often fail both questions. A line like "Your calendar is lying to you" is fun to write, but it doesn't say who it's for, what they'll learn, or why they should trust you. It gets a laugh and a scroll-past.

The fix is to swap one subjective question ("which do I like?") for five specific ones the viewer is asking without thinking.

The five checks#

Score each hook 0, 1, or 2 on each check. The maximum is 10.

1. Specificity — is there a concrete detail?#

0: Abstract ("Want to be more productive?") 1: Some detail ("A better way to plan your week") 2: A number, name, object, or situation the viewer can picture ("I plan my whole week in 20 minutes on Sunday night")

Specific hooks feel real. Vague ones feel like ads.

2. Tension — is there a gap the viewer wants closed?#

0: No question raised; the hook resolves itself 1: Mild curiosity 2: A clear gap: a contradiction, a mistake, a surprising result, or a stake ("I stopped using to-do lists and got more done")

Tension is what stops the scroll. Without it, even a specific hook is just a statement.

3. Promise — does the viewer know what they'll get?#

0: No idea what the post delivers 1: Implied payoff 2: Explicit payoff ("…here's the 3-question system")

Tension without promise is clickbait. Promise without tension is a table of contents. You want both.

4. Audience match — would your target viewer recognize themselves?#

0: Could be for anyone 1: Broad category ("creators") 2: The viewer sees their own situation ("If you freelance and every week feels like an emergency…")

This check is where most hooks lose points, and it's the one that matters most for building a real audience rather than random reach.

5. Proof — is there a reason to believe it?#

0: Pure claim 1: Implied experience 2: Evidence in the line itself: a result, a timeframe, a credential relevant to the claim, or a visible demonstration ("After 3 years of missed deadlines, this is what fixed it")

Proof is the check creators skip most, and the one that separates hooks people trust from hooks people doubt.

The process: 10 in, 2 out#

Step 1 — Write 10 hooks fast (5 minutes). Don't edit. Use different types so you're not writing ten versions of the same line: a number, a mistake, a contrarian claim, a question, a before/after, a "you" statement naming the viewer, a story opener, a result-first line, a myth-bust, and a direct how-to.

Step 2 — Score all 10 (5 minutes). Use the five checks. Be strict; a 2 means the hook clearly does it, not that it kind of could.

Step 3 — Run the delivery check (2 minutes). For the top four, ask: does the post actually deliver what this hook promises, in the first part of the content? If not, cut the hook, no matter its score. A hook that overpromises wins the first second and loses the next ten.

Step 4 — Keep the top 2. Ties go to the hook with the higher audience-match score.

Worked example: ten hooks, one topic#

Here's a made-up example. A creator who teaches productivity to freelancers is posting a short video about planning the week on Sunday night. Ten drafts, scored:

The survivors: #7 (9/10) and #9 (8/10). Both pass the delivery check, since the video shows the exact plan and mentions the lost client.

Notice what lost:

• #1, the cleverest line, scored 2. It has tension but no detail, no promise, no audience, and no proof. It's the hook the creator would probably have picked by feel.
• #8, the contrarian take, also scored 2. Picking a fight with "gurus" creates tension but gives the viewer nothing to hold onto.
• #6 scored a respectable 7 but lacked proof. It's a strong backup and a good candidate for a carousel title, where a list format suits it.

The two winners are less clever and more useful. That's the pattern you'll see almost every time.

What to do with the two survivors#

You have two strong hooks. Don't throw one away:

• Split the jobs. Use one as the spoken or written opening line and the other as the on-screen text, thumbnail text, or title. In short video, viewers often read the text overlay before the audio registers, so two strong signals at once work well together.
• Test across posts. Use hook A this week and hook B on a similar post next week. Or publish the same idea on two platforms with different hooks.
• Ask your audience. Put both in a poll or a Story question. It's not a controlled test, but it tells you what your followers respond to.

Be honest about what a single comparison tells you. One post beating another doesn't prove the hook caused it: timing, topic, and the feed that day all vary. Look for patterns across several posts. If audience-match hooks consistently hold viewers longer in your retention data, that's a signal. One win isn't.

Platform adjustments#

The five checks apply everywhere, but the hook's container changes.

Short video (Reels, Shorts, TikTok). The hook is what's seen and heard in the first seconds: the opening line, the text on screen, and the first visual. Score them together. A strong line over a static talking head loses to the same line over a visible result, because the visual is doing the proof work. Remove any intro, greeting, or logo before the hook. If you're stuck for topics to test this on, our Reels ideas and Shorts ideas lists are a fast starting point.

LinkedIn. The hook is the first two or three lines, before the "see more" cut. Treat them as one unit and score them together. Some LinkedIn hook examples that pass the five checks:

• "I lost a $12,000 client because of one email. Here's the email, and what I'd send now."
• "Freelance designers: if clients keep asking for 'one more round', your proposal is the problem. 3 lines to add."
• "We cut our weekly meeting from 60 minutes to 15. Output went up. Here's the agenda."

Each one has a number, a gap, a payoff, a named reader, and a reason to believe it, all before the fold.

Text-first posts (X, Threads, Bluesky). The hook is often the whole post or its first sentence. Promise and proof get compressed, so specificity carries more weight. Cut any word that isn't doing one of the five jobs.

Five common mistakes#

1. Overpromising. A hook that scores 10 but the post can't deliver on is worse than a 6 it can. Viewers who feel tricked leave quickly, and that early drop-off hurts the post.
2. Writing ten versions of the same hook. If all ten are "X ways to Y," you're testing wording, not approaches. Use different hook types.
3. Skipping the audience check. Broad hooks get broad, shallow reach. The audience-match score is what builds followers who stay.
4. Hooks that don't sound like you. A formula-perfect line in a voice your audience doesn't recognize breaks trust. If the winner sounds like someone else, rewrite it in your words and re-score. Our article on how "authenticity" gets misused explains why sounding like a template costs more than it gains.
5. Declaring a winner from one post. Treat each comparison as a data point, not a verdict.

Your action plan#

Before your next post:

1. Write 10 hooks in 5 minutes using ten different types.
2. Score all 10 on specificity, tension, promise, audience match, and proof (0–2 each).
3. Run the delivery check on the top four and cut any your content can't back up.
4. Keep 2. Use one as the opener and one as the on-screen text or title.
5. Log which hook you used next to the post's retention and saves. After ten posts, review which checks your best performers scored highest on. That's your audience telling you what they need from a hook.

The fastest way to get ten varied drafts is to stop writing them all yourself. Generate hook variations in TINS HUB for an idea you're already planning, then score them with the five checks instead of accepting the first one that sounds good. The tool gives you range; the scorecard gives you judgment.