EXPERIMENT 2026-04-09 · 5 min read

The Lazy Marketer's Swipe File: How to Steal Ideas Without Feeling Guilty

The Lazy Marketer's Swipe File: How to Steal Ideas Without Feeling Guilty

# The Lazy Marketer's Swipe File: How to Steal Ideas Without Feeling Guilty

I'll be honest with you: I haven't written a cold-start ad from scratch in months.

Not because I can't. Because it's stupid. There are literally millions of ads running on Meta right now, most of them paid for by someone else's budget, and a good chunk of them are actually working. Why would I sit in front of a blank doc trying to channel the muse when I could just look at what's already printing money and remix it?

That's what a swipe file is. And if you don't have one, you're doing marketing on hard mode for no reason.

What a Swipe File Actually Is

A swipe file is a folder of stuff that works. Ads, emails, landing pages, hooks, subject lines, CTAs — anything that made you stop scrolling or made you click. You save it, tag it, and come back to it the next time you need to write something.

That's it. It's not a secret. Every copywriter who's ever gotten paid has one. David Ogilvy had one. Gary Halbert had one. The difference now is that the raw material is basically infinite and free — Meta Ad Library alone has every active ad from every brand on the platform.

The problem isn't access. It's that most people don't bother.

The "Guilty" Part

Every time I tell someone to build a swipe file, at least one person goes "isn't that just copying?"

No. Copying is taking someone else's ad word-for-word and running it. That's lazy and it also doesn't work, because the ad was written for their audience, their product, their voice. A swipe file isn't about copying — it's about pattern recognition.

When I see the same hook structure show up in 12 different ads across 6 different niches, that's not coincidence. That's a pattern. Patterns are what you steal. Execution is yours.

Nobody owns "before/after split screen." Nobody owns "POV: you just realized…" Nobody owns the comment-trigger CTA. These are formats. Formats are the public domain of marketing.

If you're still feeling guilty, here's the reframe I use: every ad I save to my swipe file is an ad whose creator already paid to test it. They ran the A/B test. They burned the budget. They figured out what works. My "theft" is just… reading the research.

What to Actually Save

This is where most people mess it up. They either save nothing, or they save everything and end up with 4,000 screenshots in a folder they never open again.

Here's my rule: if it made you stop, save it. That's the only filter. Your attention is the signal. If an ad made you pause mid-scroll — even for half a second — something about it is working. Save it before you forget why.

Specifically, I save:

I don't bother saving things that look like the default. If it's a product shot with a discount badge, I already know that ad exists. I don't need another copy of it.

The Part Everyone Skips: Actually Using It

Here's the dirty secret. A swipe file only works if you look at it. And most people's swipe files are graveyards.

The way I use mine: before I write anything — a new ad, a landing page, a cold email — I spend 10 minutes scrolling my swipe file first. Not researching. Not taking notes. Just looking. I let my brain soak in the patterns. By the time I open a blank doc, I'm not starting from zero anymore. I've got 40 examples of what great looks like sitting in my head.

That's the whole trick. Input before output. You write better when you've just read a bunch of good stuff.

I also tag things aggressively. A screenshot with no tag is useless — you'll never find it again. Every ad I save gets tagged with the format (split-screen, talking-head, UGC), the hook type (curiosity, fear, aspiration), and the niche. That way when I'm writing a supplement ad and I need a good curiosity hook, I can pull up 30 of them in 2 seconds.

Why I Had Ari Build Us a Real Tool For This

I was doing all of this manually for years. Screenshots in Dropbox. Saved posts in Instagram. A Notion database I updated maybe once. It was a mess, and worse, I never looked at it because finding anything took too long.

So when I started working with Ari full time, one of the first things I told him to build was a better swipe file. Not because I wanted to sell it — I wanted to use it. I was the target customer.

That became SwipeBase. Forward any email to a custom address, it shows up in your swipe file. Tap a button on your phone, the ad you just saw gets saved. AI auto-tags everything so you don't have to. Every item is searchable. Every item has the full creative, the copy, the landing page, everything.

The whole point was to lower the friction on "save this" and "find that" to basically zero. Because the thing that kills most swipe files isn't laziness — it's friction. Every extra step between "I saw a good ad" and "it's in my file" is a step where people quit.

I use SwipeBase every day now. I have something like 600 items in mine. Before I write anything, I pop it open, scroll for 10 minutes, and then go write. My ads are better. My landing pages are better. My blog posts — including this one — are better. I'm not smarter. I'm just looking at more good stuff than I was before.

The Lazy Marketer's Cheat Code

Here's what I want you to take from this post:

Good marketers steal patterns. Great marketers have a system for stealing patterns faster than anyone else. And the ones who don't have a swipe file at all are the ones staring at blank Google Docs wondering why they're stuck.

Build the file. Save what stops you. Tag it. Look at it before you write. That's the whole playbook.

If you want the shortcut, SwipeBase is the tool I had Ari build so I could stop losing screenshots. But honestly? Even a Notion database beats nothing. Start there if you have to. Just start.

The worst swipe file is the one that doesn't exist.

---

I'm Corey. I'm building companies with an AI agent named Ari. If you want to follow along — the wins, the failures, the $173 ad campaigns that flop — subscribe to the newsletter. One email a week, no fluff.

Get the playbook. Every experiment. Every number.

I send one email per week breaking down exactly what my AI co-founder built, what it earned, and what failed spectacularly.

Subscribe Free →