If you want to know how to build a swipe file of ad hooks, the short answer is: capture every hook that stops a scroll (yours or a competitor's), tag it by type and format, store it somewhere searchable, and review it every time you write a new ad instead of starting from a blank page. The real work is in the system you build around that habit, not in the collecting itself.
What a Swipe File of Ad Hooks Actually Is
A swipe file is a personal reference library of hooks you've collected because they worked, or because they made you stop scrolling even when you had no intention of buying anything. It is not a mood board and it is not a list of ads you admire. It's a working document you pull from every time you sit down to write a new script, the same way a copywriter keeps a file of subject lines that got opened.
Most people start a swipe file the lazy way: a folder of screenshots on their phone, or a Notes app with half-finished thoughts. That version dies within a month because nothing in it is searchable or tagged. If you're going to put in the effort, build something you'll actually use when you're staring at a blank script three weeks from now.
How to Build a Swipe File of Ad Hooks: Step-by-Step
Here is the process, in order. Each step matters more than it looks because skipping one is usually why swipe files get abandoned.
1. Pick a tool that supports tags and search, not just storage
A spreadsheet beats a screenshot folder every time, because you can filter and sort. Use a simple tool like Google Sheets, Airtable, or Notion. The specific tool matters less than whether it lets you tag a hook by category and pull up every "curiosity gap" hook in one click when you need one.
2. Set up columns before you add a single hook
Build the structure first so you're not retrofitting it after you have fifty entries. At minimum, track:
- Hook text (the exact line, word for word)
- Hook type (question, bold claim, pattern interrupt, relatable complaint, stat-style, direct offer, etc.)
- Delivery format (on-screen text, voiceover line, visual gag with no words, a demo moment)
- Source (your own ad, a competitor, an organic creator, a review, a customer DM)
- Product category it was used for
- Why it worked, in one sentence (your guess, or confirmed if you tested it)
- Date added
If you want a ready-made structure instead of building columns from scratch, the hook swipe file template gives you a copy-paste starting point with these fields already set up.
3. Source hooks from four places, on a schedule
Don't wait for inspiration to strike. Set a recurring 20-minute block, weekly or biweekly, to actively hunt in these four places:
- Your own top-performing ads. Pull the exact opening line from any ad that has outperformed your account average. This is your highest-value source because it's proven in your own audience.
- Competitor and adjacent-category ads. Scroll ad libraries for your category and one or two adjacent categories. A supplement hook can often be reworked for a skincare product, and a software hook can be rewritten for a service business.
- Organic content in your niche. Creators who get strong engagement without paying for it are doing hook work for free. The opening line of a well-performing organic video is often a hook you can adapt.
- Your own customer language. Reviews, support tickets, and DMs are full of phrases customers use to describe their problem before they found your product. Those phrases make some of the most honest, highest-converting hooks because they're not copywriter language, they're the customer's own words.
4. Log the hook raw, then tag it
Write down the hook exactly as it appeared, with no editing. Editing can wait. The point of the raw capture is to preserve the actual rhythm and word choice that worked, because small wording changes can quietly kill what made a line effective in the first place. Tag it immediately after, while the context is still fresh in your head.
5. Review the whole file before writing anything new
Before you write a script for a new ad, open the swipe file and filter by hook type or by product category first. Read through ten to fifteen entries before you write a single new line. This step is what separates a swipe file from a graveyard of screenshots: it only earns its place if you actually open it before writing instead of after you're stuck.
Categorizing Hooks So You Can Actually Find Them Later
Tagging by type is what makes a swipe file useful instead of just long. A simple starting taxonomy, which you can expand as your file grows:
| Hook type | What it does | Example pattern |
|---|---|---|
| Pattern interrupt | Breaks a visual or verbal expectation in the first second | Someone throws the product, then catches it mid-sentence |
| Direct question | Asks the viewer something they can't help but answer in their head | "Why does your skin still look dull after three serums?" |
| Bold claim | States a strong outcome up front, to be proven later in the ad | "This fixed a problem my dermatologist couldn't." |
| Relatable complaint | Names a frustration the viewer has felt but maybe not voiced | "I was tired of charging four different cables every night." |
| Myth-bust / correction | Challenges a common belief the viewer holds | "You've been storing your coffee wrong this whole time." |
| Social proof opener | Leads with a result, review, or reaction instead of a claim | "My partner asked what I did differently in two weeks." |
If you tag every entry with one of these categories (plus any new ones you discover), you can filter your swipe file to "show me every relatable complaint hook" when you're stuck on a specific ad. That's a far more useful search than scrolling a screenshot folder hoping something jumps out.
Turning a Swipe File Entry Into a Usable Hook
A swiped hook is a starting point, not a script you copy word for word. The goal is to borrow the structure and rhythm, then rewrite it around your own product, audience, and voice. A simple three-step adaptation process:
- Identify what made the hook work mechanically: is it the question structure, the specificity of the number, the visual surprise, or the emotional nerve it hits?
- Strip out anything specific to the original product or brand and replace it with your own product's real detail. Vague swaps produce vague hooks.
- Say the new version out loud. If it sounds like something a real person would actually say, it's probably usable. If it sounds like ad copy, rewrite it again.
For a deeper breakdown of writing strong openers from scratch, see how to write a hook in the first three seconds. And since one script almost never needs just one hook option, how to write multiple hooks for one ad walks through generating several angles from a single swipe file entry.
Keeping the Swipe File Alive Instead of Letting It Go Stale
Most swipe files die from neglect, not from a bad setup. A few habits keep it useful long-term:
- Set a recurring calendar reminder to add new hooks, weekly or biweekly, rather than relying on remembering.
- Mark hooks as tested, winning, or retired, so you know at a glance what's proven versus still speculative.
- Archive hooks that have been run heavily and are clearly fatigued in your account, so the top of the file always shows your freshest options.
- Revisit old, unused entries every few months. A hook that didn't fit last quarter's offer might be exactly right for a new product line or seasonal promo.
If you're running several hooks in parallel to see which one actually performs, pair your swipe file habit with a structured testing process. How to test multiple ad hooks simultaneously and how many hooks to test per ad campaign both cover how to turn a shortlist of swiped-and-adapted hooks into an actual test plan without wasting budget.
Where FrameNotion Fits Into the Process
A swipe file solves the writing problem. It doesn't solve the production problem, which is usually the bigger bottleneck: even with ten strong hook options logged and ready, turning each one into a finished video ad still takes real time if you're filming, editing, and captioning by hand.
FrameNotion is built for the second half of that workflow. You paste a product or website link, and FrameNotion AI writes a custom 30-second vertical ad from scratch, covering hook, problem, benefit, proof, offer, and call to action, then checks its own rendered frames and renders the MP4. If you've already got a hook from your swipe file that you want to test, you can drop it into the notes field along with your offer details, and the ad gets built around it. A finished ad takes about 10 to 20 minutes, and the first one is free on sign-up, so you can see how a swiped-and-adapted hook performs in finished video form before committing to a plan. See how it works or browse example ads to get a sense of the output.
Worth noting: FrameNotion doesn't publish ads to ad platforms or track their performance after launch. It handles the writing-to-finished-video step; you still run and monitor the campaign yourself.
Once you've got a video ad live, your swipe file habit also helps you spot when it's time to retire a hook. If engagement on a specific hook starts dropping, these signs an ad creative is underperforming are worth checking before you assume the hook itself is the problem rather than the audience, offer, or placement.
Frequently asked questions
How many hooks should be in a swipe file before it's actually useful?+
There's no magic number, but somewhere around 30 to 50 tagged entries is usually enough to start seeing patterns across hook types. The value comes from the tagging and search structure more than raw volume, so a well-organized file of 30 hooks beats a messy folder of 300 screenshots.
Should I only swipe hooks from my own industry?+
No. Some of the most useful adaptations come from adjacent or unrelated categories, because a strong hook structure (a question, a myth-bust, a relatable complaint) usually transfers across products even when the specific wording doesn't. Collect broadly and tag by hook type so you can filter later regardless of the original industry.
Is it legal or ethical to swipe a competitor's hook?+
Swiping the structure and rhythm of a hook and rewriting it around your own product is standard creative practice. Copying a hook word for word, especially if it includes a specific claim or branded phrase, is both a weak move creatively and a potential problem legally. Treat swiped hooks as inspiration to adapt, not text to copy-paste.
How often should I update my swipe file?+
Set a recurring weekly or biweekly block to add new entries, and revisit the whole file every few months to retire hooks that have gone stale or been overused in your own campaigns. A swipe file that isn't updated regularly slowly turns back into the screenshot folder it was meant to replace.
Can a swipe file help with video ads specifically, not just written copy?+
Yes. Tag each entry with its delivery format (on-screen text, spoken voiceover line, or a purely visual pattern interrupt) so you know which hooks need to be read, heard, or simply seen on screen. That distinction matters for video ads, since a hook that reads well as text might land differently as spoken voiceover.