A good hook does one job in the first three seconds: it names a problem the viewer has, opens a question, shows proof, makes an offer or picks a fight with a belief. Use problem and curiosity hooks on cold audiences and proof and offer hooks on warm ones. Test 3 to 5 openers on the same ad body, then keep the one with the lowest cost per result.
Coming up with hooks takes an afternoon. Knowing which angle is winning is harder: you have to read results across hundreds of ads and several networks, not just spot the one ad that got lucky. Segwise is an AI creative analytics and creative generation platform for that problem: it tags the hook line, on-screen text, opening visual and offer in every ad you run on Meta, TikTok and your other networks, then joins those tags to each network's results and your MMP events. Through the Segwise MCP in Claude or ChatGPT, you can ask "which hook angle brought our cheapest installs last month?" and get the answer from your own tagged ads, then make the next batch of hooks from that winning angle in Segwise.
TL;DR
- DTC brands: start with problem hooks (1 to 20) and proof hooks (41 to 60). Most DTC ads are statics, so treat the image and headline as the hook too.
- Apps: lead with a daily-life problem or a curiosity hook that shows the app in use. Offer hooks are a second-line angle for apps; save them for trial and annual-plan pushes.
- Mobile games: challenge, cliffhanger and reverse-psychology hooks (22, 33, 81) fit cold installs. Real gameplay and player reactions carry proof.
- Measuring: hook rate is 3-second plays divided by impressions. One 2026 benchmark puts the median at 28% on Meta and 33% on TikTok.
- Testing: change only the opener, run Meta's creative test with 2 to 7 copies, and judge on cost per result, not on hook rate alone.
What makes a hook idea worth testing?
A hook is the opening of an ad: the first frame, the first line (spoken or on screen) and the first motion. On video it's roughly the first three seconds. On a static ad it's the image and headline the eye lands on first.
An idea is worth a test slot when it passes three checks:
- It works with the sound off. Put the hook line on screen as text, not only in the voiceover.
- It points at the ad's payoff. If the hook promises a fix for wet socks, the next ten seconds show dry socks. A hook that the body can't pay off raises hook rate and wastes the click.
- It's one idea. A problem, or a question, or a price. Stacking three angles in three seconds reads as noise.
Every idea below is written as a line you could say or put on screen, with a note on who it fits and when to use it. Swap in your own product, and keep any price, deadline or guarantee true.
100 hook ideas, sorted by angle
The five angles map to how far along your viewer is. Problem and curiosity hooks stop people who have never heard of you. Proof and offer hooks convert people who already know you. Contrarian hooks work in crowded categories where every ad sounds the same.
Problem hooks (1 to 20): name the pain the viewer already feels
| # | Hook idea | Fits | Use it when |
|---|---|---|---|
| 1 | "If your concealer creases by lunch, it's not your skin." | DTC | The pain is visible and buyers blame themselves for it |
| 2 | "You pay for four streaming apps and watch one." | Apps | Money leaks quietly and the viewer hasn't added it up |
| 3 | "Stuck on level 40? Everyone is. Here's why." | Games | A difficulty spike is a shared frustration in your genre |
| 4 | "Stop rinsing your shaker bottle with cold water." | DTC | A small daily habit causes a smell or mess people recognise |
| 5 | "Phone on 20% by 3 p.m. again?" | Apps, DTC | The problem shows up at the same time every day |
| 6 | "This is what an eight-year-old mattress looks like inside." | DTC | Your product replaces something people keep too long |
| 7 | "Three months of Spanish lessons and you still freeze ordering coffee." | Apps | Viewers have tried another method and stalled |
| 8 | "Your dog isn't bad. He's bored." | DTC | Buyers misread the cause, and your product fixes the real one |
| 9 | "Every budget app made me feel worse about money." | Apps | Your category has a known emotional downside you avoid |
| 10 | "Wet socks by 9 a.m. in waterproof boots?" | DTC | A rival's promise fails a real-world test |
| 11 | "Your kid asks for the tablet the second you sit down." | Apps, Games | You sell to parents and the moment happens daily |
| 12 | "You lose your keys twice a week. Here's the fix." | DTC | Frequency makes the problem concrete |
| 13 | "Still awake at 3 a.m. scrolling? Same." | Apps | The viewer is likely watching at the moment the problem happens |
| 14 | "If your cold brew tastes sour, check the grind." | DTC | You can teach something useful before you sell |
| 15 | "Your base gets raided while you sleep." | Games | Loss aversion is part of the game loop |
| 16 | "Most people store tomatoes in the fridge. That's the mistake." | DTC | A common mistake has a fix that is your product |
| 17 | "Twenty minutes deciding what to cook, again?" | Apps | Decision fatigue is the pain, and evening placements are cheap |
| 18 | "Your gym bag smells before you open it." | DTC | The problem has a strong visual you can show with no sound |
| 19 | "Bored of puzzle games that all play the same?" | Games | You target players who already install your genre |
| 20 | "Your store gets visits but no sales." | Apps | You sell to business owners who track a clear number |
Curiosity hooks (21 to 40): open a question the ad answers
| # | Hook line | Best for | When to use it |
|---|---|---|---|
| 21 | "I didn't expect the second one to win." | Any | You can show the reveal by second five |
| 22 | "Guys, I think I reached the last level." | Games | Your game has long progression players wonder about |
| 23 | "The secret ingredient is already in your pantry." | DTC | The reveal is the product's key component |
| 24 | "Watch what happens when she taps the red one." | Games, Apps | The first action in the app or game is satisfying to watch |
| 25 | "Nobody at the store tells you this about running shoes." | DTC | The purchase is high-consideration and buyers want insider advice |
| 26 | "I changed one setting. Look at my screen time now." | Apps | One feature delivers a visible before and after |
| 27 | "Why do hotel pillows feel like that?" | DTC | You can ask about an experience almost everyone has had |
| 28 | "Wait for the last frame." | Any | The ad runs under 15 seconds and the payoff is visual |
| 29 | "My dentist asked what I changed." | DTC | A third party noticing the result is more believable than you saying it |
| 30 | "There's a reason chefs don't use nonstick for this." | DTC | Expert habits explain why your product is built differently |
| 31 | "What happens if you practice one minute a day for a month?" | Apps | Small daily effort is the product's promise |
| 32 | "Can you spot the fake?" | Any | You sell dupes, design or a spot-the-difference game |
| 33 | "Most players miss the move in this puzzle. Do you see it?" | Games | Your game rewards a clever solve |
| 34 | "The packaging is the best part. Hear me out." | DTC | Unboxing and gifting season are near |
| 35 | "I asked strangers on the street which one smells better." | DTC | Scent, taste or feel can be judged on camera |
| 36 | "Here's what a week of groceries looks like with this app." | Apps | The result is a physical thing you can show |
| 37 | "POV: you finally know why your plants keep dying." | DTC, Apps | The POV format is native on TikTok and Reels |
| 38 | "Don't skip to the end." | Any | You pair it with a timer or progress bar on screen |
| 39 | "Rate my desk setup out of ten." | DTC | You want comments, and the product is part of a setup |
| 40 | "I ran this level 50 times to find the trick." | Games | Visible effort signals a payoff worth waiting for |
Proof hooks (41 to 60): show it works before you explain it
| # | Opener | Audience | Works best when |
|---|---|---|---|
| 41 | Silent review: no voice, just a player's face reacting to the game | Games | The reaction is genuine and the ad must work muted |
| 42 | "Reading our one-star reviews out loud." | DTC | Your brand can turn doubt into trust with humour |
| 43 | Split-screen before and after in the very first frame, no words | DTC | The change is visible in one frame (check your category's claim rules) |
| 44 | "Replying to your comment: yes, it holds a full mug." | DTC | One objection keeps showing up in comments |
| 45 | "My husband said it wouldn't fit. It fit." | DTC | A sceptic converting is the story |
| 46 | "My grocery bill before and after I started meal planning." | Apps | You can show a real receipt or screen |
| 47 | Pour test: your product next to the market leader, same cup | DTC | A head-to-head demo settles the claim in seconds |
| 48 | "A 30-day streak, screenshots, no edits." | Apps | Retention or habit is the proof point |
| 49 | "A nurse on a 12-hour shift tests our insoles." | DTC | A job makes the pain obvious and the test credible |
| 50 | Rapid-fire Q&A: players ask, the developer answers | Games | You run live ops and have a community asking questions |
| 51 | "We sent it to our harshest customer." | DTC | Your brand voice is self-aware |
| 52 | Customer-filmed unboxing, first frame on the box at the door | DTC | Delivery speed or packaging is a selling point |
| 53 | "I tested five budgeting apps for a month. This is the one I kept." | Apps | Buyers compare options before they commit |
| 54 | Stress test: drop it off a roof, then use it | DTC | Durability is your edge and it films well |
| 55 | Screen recording of a real win, with the player's voice: "No way." | Games | Real gameplay shows players exactly what they install |
| 56 | "Customer since 2019. This was her first order." | DTC | Repeat purchase is the proof, not a single review |
| 57 | "Our support team reads the strangest message we got this month." | DTC, Apps | Human service is part of what people pay for |
| 58 | "My trainer made me try this." | DTC, Apps | An expert's nudge lends borrowed authority |
| 59 | Reviews scrolling on screen while the product works in the background | DTC | You retarget viewers who already know the brand |
| 60 | "Watch me beat this level with zero boosts." | Games | Players suspect the game is pay-to-win |
For more on reviews, comments and creator videos as proof, see our guide to UGC, testimonials and comments as social proof.
Offer hooks (61 to 80): lead with a price, a deal or a deadline
Only run these with a real price, a real deadline and a guarantee you honour. A fake countdown gets clicks once and refunds after.
| # | Hook idea | Best for | Run it when |
|---|---|---|---|
| 61 | "This $40 serum replaced my $120 one." | DTC | You're the value pick against a premium leader |
| 62 | "Free shipping ends at midnight." | DTC | The deadline is real and shipping cost is the objection |
| 63 | "First lesson free. No card needed." | Apps | Sign-up friction is what stops installs |
| 64 | "Back in stock. It sold out in a weekend last time." | DTC | A true restock follows a real sell-out |
| 65 | "What a dollar a day gets you." | Apps | An annual price looks big until you divide it |
| 66 | "Build your own bundle and save on every extra item." | DTC | Average order value is the goal |
| 67 | "Starter pack: gems and a rare skin for new players." | Games | You run a new-player offer on install campaigns |
| 68 | "Sleep on it for 100 nights. Hate it? Send it back." | DTC | Risk of a wrong purchase is the biggest objection |
| 69 | "A gift under $30 that doesn't look under $30." | DTC | Gifting season is on and the product looks premium |
| 70 | "Your first box is half price." | DTC | Subscription brands need the first order |
| 71 | "We cut out the middleman. Here's the price difference." | DTC | Your margin story is the reason you're cheaper |
| 72 | "Annual plan math: twelve months for the price of seven." | Apps | The annual discount is real and easy to show |
| 73 | "The Halloween map closes Sunday." | Games | A live event has a real end date |
| 74 | "Price drop on the bundle everyone asked for." | DTC | You have proof people asked for it, such as comments |
| 75 | "The refill costs less than your coffee." | DTC | Replenishment and refills drive repeat revenue |
| 76 | "Pick any three for the price of two." | DTC | You want to move several products at once |
| 77 | "Log in seven days in a row and the legendary car is yours." | Games | Day-7 retention matters more than the install |
| 78 | "Student? Show this ad at checkout." | DTC, Apps | One audience gets its own discount |
| 79 | "We'll cover your switching fee." | Apps | Switching cost keeps people with a rival |
| 80 | "Trade in your old one for credit." | DTC | Buyers already own an older version |
Only run price and deadline hooks when the offer is real; a fake deadline costs trust on the next ad too.
Which angle is winning in your account?
A list can't tell you that. The same problem hook can win for one skincare brand and flop for another, because the audience, offer and ad body differ. What you can know is which angle keeps winning in your own account once every ad is tagged by its hook. The panel below uses the impressions-based hook rate from this post.

See which hook angle wins: Segwise lets you tag every ad's opening line and first frame by angle, join those tags to each network's hook rate, spend and ROAS plus your MMP events, and settle which angle gets next month's budget. Book a demo.
Contrarian hooks (81 to 100): disagree with something your viewer believes
| # | Hook line | Fits | Reach for it when |
|---|---|---|---|
| 81 | "Don't download this game if you hate winning streaks." | Games | Reverse psychology fits a confident, playful brand |
| 82 | "You don't need a ten-step skincare routine." | DTC | Your category overwhelms buyers with steps |
| 83 | "Most protein bars are candy bars with better marketing." | DTC | Your product is the cleaner alternative |
| 84 | "I hate meditation apps. This is a breathing game." | Apps | Your app solves the problem in a different format |
| 85 | "Don't buy this if you only cook on weekends." | DTC | Qualifying buyers cuts returns and lifts trust |
| 86 | "Unpopular opinion: your workout is too long." | Apps | Short sessions are your product's point |
| 87 | "Our worst-selling colour, and why we keep making it." | DTC | Brand story and values drive the purchase |
| 88 | "The tutorial is the hardest level. Sorry." | Games | Your game is genuinely challenging and proud of it |
| 89 | "Expensive running shoes aren't making you faster." | DTC | You're the value brand in a premium category |
| 90 | "Delete your to-do list." | Apps | Your app replaces lists with a different system |
| 91 | "This ad is for people who gave up on language apps." | Apps | Lapsed users of the category are your best audience |
| 92 | "We're not the cheapest. Here's why that matters." | DTC | Price is the objection and quality is your answer |
| 93 | "Don't play this before bed." | Games | The game is easy to lose an hour in |
| 94 | "Fast fashion is cheaper. Until you count the washes." | DTC | Cost per use beats sticker price |
| 95 | "Most budgeting advice is written for people who already have money." | Apps | Your audience feels left out by the category |
| 96 | "Pay-to-win? Watch a free player top the leaderboard." | Games | Players doubt the game is fair |
| 97 | "I'm the founder and I almost shut this down." | DTC | A founder story is true and has a turning point |
| 98 | "Your phone isn't the problem. Your home screen is." | Apps | You reframe the problem so your app is the fix |
| 99 | "We tested it against the original and lost on one thing." | DTC | Admitting one flaw makes the other claims believable |
| 100 | "Skip the gym this week. Do this at home." | Apps | Home workouts are the product and January is close |
Hook formats that work with any angle
The 100 lines above are what to say. These formats are how to show it, and each one can carry any of the five angles.
| Format | How it opens | Pairs well with |
|---|---|---|
| Handwritten on-screen text | A hand writing the hook line on paper, packaging or a mirror | Problem and proof hooks for DTC brands |
| Numbered breakdown | "3 things I wish I knew before buying a bike" with the list number on screen | Curiosity hooks, and statics as well as video |
| Behind the scenes | The product being made, packed or tested, with no polish | Proof hooks and founder stories |
| Satisfying first frame | A clean pour, a perfect cut or a smooth swipe that plays out in a second | Curiosity hooks that must work muted |
| Comment reply, stitch or duet | A real comment on screen, then the answer | Proof hooks that tackle one objection |
| Creator collab | Two creators testing the product together | Proof hooks on TikTok and Reels |
| Slideshow with a cliffhanger | Quick slides of key moments, with the reveal held for the last one | Curiosity hooks for games and apps |
| Ad fatigue callout | "Tired of seeing the same ads? Try something new." | Contrarian hooks for games and apps |
Which hook types work for DTC brands, apps and games?
Each audience buys for different reasons, so the strongest angle shifts. This table is a starting point built from the "fits" column above; your own data should overrule it within a month.
| Angle | DTC brands | Apps | Mobile games |
|---|---|---|---|
| Problem | Strong: shows the pain the product removes | Strong: a daily moment the app fixes | Medium: frustration with a genre or a level |
| Curiosity | Medium: reveals, tests, "why does" questions | Strong: show the app doing one surprising thing | Strong: challenges and cliffhangers |
| Proof | Strong: before and after, reviews, demos | Medium: streaks, screenshots, comparisons | Strong: real gameplay and player reactions |
| Offer | Strong for warm audiences and sales events | Medium: trials, annual plans, switching offers | Medium: starter packs and live events |
| Contrarian | Medium: works in crowded categories | Medium: reframes a tired category | Medium: reverse psychology and difficulty pride |
Should DTC hooks focus on statics or video?
Both, and statics deserve more attention than they get. In one analysis of 67,852 active ads from 106 brands, static images made up 64.8% of ads from DTC brands. On a static, the hook is the image plus the headline, so most of the 100 lines above work as a headline over a strong product or problem shot. For DTC-specific openers built around offers, products and buyer psychology, see our guide to DTC ad creative hooks.
How do you pick angles for this month?
A rule of thumb for teams launching 10 or more new ads a month:
- Cold prospecting: put at least half your new hooks into problem and curiosity.
- Retargeting: use proof and offer hooks, since these viewers already know what you sell.
- Spread your bets: of every 10 new hooks, write about 4 in your current best angle, 3 in your second best and 3 in an angle you haven't tested yet.
- Contrarian: add one or two when your ads start to look like everyone else's in the ad library.
If you launch fewer than 10 ads a month, test one angle at a time so each test has enough spend to read.
What's a good hook rate?
Hook rate is the share of people who watched at least three seconds of your video. On Meta, the usual formula, and the one our hook rate guide uses, is 3-second video plays divided by impressions. Meta counts a 3-second play as "the number of times that your video played for at least three seconds, or for nearly its total length if it's shorter than three seconds," and replays don't count, according to Meta's definition of 3-second video plays. TikTok Ads Manager reports 2-second and 6-second video views instead, so TikTok numbers aren't directly comparable with Meta's.
Two formulas are in use, and they give different numbers. Dividing 3-second plays by impressions measures whether people stopped scrolling; our glossary calls that thumb-stop rate. Dividing 3-second plays by video plays measures whether people who started watching kept going; our glossary calls that hook rate, and it comes out higher. Both are fine. Pick one, name it, and use it in every report.
Hook rate benchmarks for 2026
Benly's 2026 ad creative benchmarks, based on 12,347 ads across 842 ad accounts from January to March 2026, put the numbers here:
| Platform | Median | Top 25% | Top 10% |
|---|---|---|---|
| Meta (all placements) | 28% | 37% | 45% |
| TikTok | 33% | 44% | 55% |
Treat these as one vendor's dataset, not a law. As a rule of thumb, if a Meta ad sits under about 20%, rework the first frame and first line before you touch anything else. For the full playbook on lifting the number, read our guide on what hook rate is and how to improve it.
Is a high hook rate enough?
No. A hook that stops the scroll but promises the wrong thing gets you cheap views and expensive customers. Read hook rate next to hold rate and cost per result. Hold rate isn't one of the named video metrics on Meta's help pages, so you build it yourself. One way is 15-second ThruPlays divided by 3-second plays; Meta counts a 15-second ThruPlay when a video is "played to completion, or for at least 15 seconds," according to Meta's 15-second ThruPlays definition.

- Low hook rate: the opener is the problem. Change the first frame or first line.
- Good hook rate, low hold: the opener works but the middle loses people. Fix the seconds after the hook.
- Good hook and hold, high cost per result: the ad works but the offer or landing page doesn't. Don't blame the hook.
How do I test hooks?
Test hooks by changing only the opener and keeping everything else the same. If the body, end card and CTA also change, you can't tell what moved the result.

Step 1: Write 3 to 5 openers for one ad body
Take your current best ad. Write 3 to 5 new openers from different angles, for example one problem, one curiosity and one proof hook. Cut each one onto the same body, so the first three seconds are the only difference.
Step 2: Run a creative test in Meta Ads Manager
Meta's creative test lets you compare up to seven variants inside an existing campaign, and Meta says the delivery system's learnings are kept, since moving winners to another campaign would reset them. As of November 2026:
- In Ads Manager, open an existing campaign. Creative tests only run on campaigns using the Highest volume bid strategy.
- At the ad level, scroll to the Creative test section and click Set up test.
- Choose 2 to 7 copies of the ad, then swap the creative on each copy for one of your hook variants.
- Set how much budget the test ads get. Meta suggests no more than 20% of the existing budget.
- Set the test duration and the metric to compare on, such as cost per result, then publish.
Two details matter. Meta says "a confidence level is not included" in the results, and the test ads keep running after the test until you turn them off.
Step 3: Use an A/B test when you need a clean split
If you need a stricter read, use Meta's A/B testing. Each version goes to a separate slice of the audience "ensuring that nobody sees both," and Meta recommends the same budget for both versions. Meta also advises against testing by switching ads on and off by hand, because audiences overlap and results become unreliable.
Step 4: Read hook rate first, then cost per result
Hook rate needs only impressions and plays, so you can read it before cost per result, which needs conversions. As a rule of thumb, treat a hook rate gap of under 3 points between two variants as a tie, and let cost per result pick the winner.
Step 5: Tag the winner by angle, not just by ad name
A single winning ad tells you little. A pattern tells you what to write next. Label each hook with its angle (problem, curiosity, proof, offer, contrarian), so after a few rounds you can see which angle wins for each audience. Our guide on how to analyze hook patterns in video ads walks through doing this at scale.
How Segwise helps you find your winning hook angles
Tagging hooks by hand is manageable for a few dozen ads, and slow once you run hundreds across several networks. Segwise lets your team tag every opener once, join it to results from each network and your MMP, and answer "which angle should we write more of?" from data instead of rewatching ads.
Tag every opener by angle

- Creative Tagging tags hook dialogue, on-screen text, characters, offers, visuals and pacing across video, image and playable ads, with 20+ standard tags, as in the view above. Add a custom tag for hook angle (problem, curiosity, proof, offer, contrarian), or pull it from file names such as PULSE_PROOF_RestDay with nomenclature tagging. Your team reviews and adjusts the tags.
- Tag-to-metric mapping joins each tag to IPM, CTR, CPA, ROAS and custom events from 15+ ad networks and your MMP (AppsFlyer, Adjust, Branch, Singular or Kochava), so a games team can check whether challenge hooks that win on installs also hold up on D7 ROAS, and move budget to the angle that pays back. Connecting Meta is a sign-in and a few selections, as the Meta integration guide shows.
Ask, score and make new hooks
- The Creative Strategy Agent answers questions such as "Which hooks drove the most installs last month?" from your tagged ads. With asset clustering, it groups ads that share the same footage, so you can see whether swapping a problem hook for a proof hook on one ad body caused the ROAS gap.
- The Segwise MCP brings the same tagged data into Claude, ChatGPT, Cursor, Codex or Grok Bot. It is available to every Segwise account: connect your AI tool, sign in with your Segwise login, and ask "which offer hooks beat our proof hooks on cost per purchase in October?" It reads your Segwise data and answers; it can't change your campaigns or ad data.
- Pre-flight Testing lets you upload up to 10 new creatives at a time and scores each one out of 100 against the ads already winning in your account, with a run-or-hold verdict. Hook strength is one of the five things it scores, so you can drop the weakest of your five new openers before any of them spend.
- The Creative Generation Agent lets you make statics and videos up to 30 seconds, including AI UGC-style ads, built around the hook angle your tags show is winning, such as five new creator-to-camera proof openers for your best ad body.
You can use each agent on its own, or do it all in one place by chatting with the Segwise agent: ask how your ads are performing, see what competitors are running, and make new ads from your winners.
Where Segwise falls short, and when a spreadsheet will do
Segwise needs your ad accounts connected to learn what wins, and Pre-flight scores against your own winners, so a brand-new account has little to compare with until a few weeks of results come in. Tagging shows which angles go with better results; a creative test is still how you prove a single hook caused the lift. Segwise starts at $999/month for advertisers spending up to $250K/month; startups spending under $50K/month pay $499/month (limited eligibility). If you launch fewer than 10 ads a month on one network, a spreadsheet with an angle column will do. For a side-by-side of tools, see our comparison of hook tagging and analysis tools.
Which hook ideas should you test first?
Start with the angle that matches your audience: problem or curiosity for cold traffic, proof or offer for people who already know you. Pick 3 to 5 hooks from that angle in the list above, cut them onto your best-performing ad body, and run them as one creative test. Judge on cost per result, tag the winner by angle, and write next month's hooks from the angle that keeps winning.