Everyone who works with Facebook ads knows what a swipe file is. Almost nobody has one that's actually useful.
The typical version is a folder of screenshots, maybe a Notion page with some links, or a Slack channel where people drop ads they like. Three months later, nobody can find anything in it. The folder keeps growing. The links rot. The channel gets buried. And when someone needs to brief a creative, they ignore the swipe file and start from scratch.
A swipe file that doesn't get used is just a collection. What you need is a system. The reason ad teams build swipe files in the first place is to create a repeatable advantage in creative production, but that only works when the file is organized enough to actually use.
The difference between collecting and building a reference system
Collecting is passive. You see an ad you like, you screenshot it, you move on. Building a reference system requires one extra step: capture why you saved it.
That step is the difference between a folder you forget and a library you return to. An ad in your collection without context is just an image. An ad tagged with the hook type, the emotion it triggers, the offer structure, and the audience signal it carries is a lesson you can act on.
This is the core idea behind a properly built swipe file. Not a collection of things you found interesting, but a structured reference system you can search when you have a specific creative problem to solve.
What to capture when you save an ad
When you save an ad, record these five things:
- Hook: What is the first line or first frame? What pattern does it use: question, bold claim, visual surprise, identity trigger?
- Format: Video, image, or carousel. UGC or polished. Short or long form.
- Offer: What is being sold and how is it framed? Discount, urgency, free trial, transformation?
- Audience signal: Who is this for, based on the creative? What identity, problem, or aspiration does it speak to?
- Why it works: One sentence on why you saved this. What specifically does it do well?
This takes maybe 60 seconds per ad. Over time, these notes compound. When you need to brief a video that opens with a bold claim targeting people who've tried other solutions, you can search for exactly that and find five examples with notes on what made each one work.
A tagging system that's actually usable
Tags are only useful if they're consistent. A tag structure that one person creates organically will be different from what the next person creates, and you'll end up with a mess of overlapping and inconsistent labels.
Pick a taxonomy upfront and stick to it. Here's a simple one that works:
- By format:
video-ugc,video-polished,image-product,image-lifestyle,carousel - By hook type:
hook-question,hook-bold-claim,hook-identity,hook-visual,hook-problem - By emotion:
emotion-fear,emotion-desire,emotion-curiosity,emotion-humor,emotion-aspiration - By offer type:
offer-discount,offer-urgency,offer-free-trial,offer-transformation,offer-comparison - By niche or vertical:
niche-skincare,niche-supplements,niche-apparel(add as needed)
This structure lets you search across multiple dimensions at once. "Show me all UGC video ads that use a fear hook" is a useful query. "Show me everything I saved in January" is not.
Storage options: Notion, Airtable, or dedicated tools
Notion works for solo researchers or small teams. It's flexible, easy to add custom fields, and most teams already use it. The limitation is that video doesn't embed well from Facebook, links expire, and it becomes slow when your library gets large.
Airtable is better for teams that need to filter and sort across many fields. It's a better database than Notion and handles larger volumes more cleanly. The tradeoff is more setup time and a steeper learning curve for non-technical team members.
Both have the same core problem: you have to manually copy ads into them. Paste the video URL. Upload a screenshot. Add the metadata. This is fine when you're saving one or two ads. It's annoying at scale. There are also dedicated ad spy tools worth comparing if you're doing this at volume.
The better approach for teams doing regular ad research is a dedicated tool. Spreshapp's saved ads library is built specifically for this. You save ads directly from the Meta Ad Library with the Chrome extension in one click. The ad creative, metadata, and advertiser info are all captured automatically. Then you add your own tags and notes. No copy-pasting, no expired links, no screenshots.
How Spreshapp acts as a live swipe file
A static swipe file gets stale. The ads you saved six months ago were relevant then. What's running today is what you need to understand now.
Spreshapp solves this by connecting your swipe file directly to live ad data. With the ad browser, you can search by keyword or competitor to see what's running right now, filter by how long ads have been live, and save the ones worth studying.
An ad that has been live for 60 days isn't there by accident. Brands don't pay to keep ads running that aren't converting. This gives you a filter that no screenshot folder can replicate: real market validation. You're not saving ads because they looked good to you. You're saving ads because a brand decided they were worth spending money on for two months.
That's a different kind of swipe file. It's a live read on what your market is responding to, organized around your own tags and research.
How to use your swipe file when briefing creative
The swipe file is useless if it sits there between briefs. The way to make it valuable is to open it at the start of every brief, not the end.
Before you write a brief, run a search: what do we have tagged with this hook type or this emotion? Pull three to five examples. Include them in the brief as reference, with a note on what to replicate and what to do differently.
Creators and copywriters work better with visual references than written descriptions. "Do something like this, but for our audience" with a specific example is more useful than three paragraphs describing a tone. The swipe file makes this possible.
It also prevents briefs from getting worse over time. Without references, briefs drift toward vagueness. With examples, the bar is set by something concrete.
Keeping it fresh: how often to add, when to prune
Add to your swipe file consistently, not in bursts. Fifteen minutes a week of focused ad research, run through Spreshapp or the Meta Ad Library, is more valuable than a two-hour binge every few months. The weekly habit keeps your reference points current. If you're not sure where to look, the guide on how to find winning ads on Facebook covers the most reliable sources and filters.
Prune once a quarter. Remove ads from formats or niches you no longer work in. Remove anything where the link is dead and you don't have a saved copy. Remove ads you saved because they were "interesting" but that you've never actually referenced. A smaller, denser library is more useful than a large one full of noise.
The goal isn't the biggest swipe file. It's the most useful one. That means every ad in it should be there for a reason you can still articulate.
If you want to see how a well-structured saved ad library actually works in practice, the saved ads section in Spreshapp is set up for exactly this. Tags, notes, folders, search, all connected to live ad data so your reference library never goes stale.