Build a swipe file

Advertising legends once collected folders full of ads they had ‘swiped’ a copy of. They’d consult the file when writer’s block struck, and they were in need of inspiration. Now swipe files are mostly found in note-taking apps, but the principles remain the same. Memory is a poor storage format. We’re hardwired towards novelty, so the even great ideas we’ve witnessed get pushed out. Building a second brain, a digital one that never forgets, means you’ll never forget a good idea.

In my last company we were famous for building the “Playbook”: a database of 800+ proven marketing tactics – ad copy, design styles, targeting methods – used by tens of thousands of marketers. It started as an Excel spreadsheet of things I had tried that worked. Rather than reinvent the wheel every time, I would scan through the file whenever we got a new client. We’d always hit the ground running with a list of winning ideas, and whenever we got stuck, we knew where to look for inspiration. One good idea could be applied across multiple clients.

Swipe files serve as a reminder of greatness. Jack London and Robert Louis Stevenson (Dr. Jekyll and Mr. Hyde) were famous for using copywork – copying the works of masters – to learn how to write fiction. Ben Franklin spent years doing the same to learn the newspaper business and the art of persuasion. Comedy director Judd Apatow (The 40-Year Old Virgin, Knocked Up, This is 40, Funny People, and dozens of others) used Saturday Night Live scripts for copywork to learn the structure of a good joke. Hunter S. Thompson typed out The Great Gatsby word for word because he “wanted to know what it felt like to write a masterpiece”.

If we’re the average of the five people we spend the most time with, then the writer who spends their time surrounded by the works of F. Scott Fitzgerald, Ernest Hemingway, Jack Kerouac, George Orwell, and Henry Miller – as Thompson did – has a clear advantage over someone reading whatever’s currently fashionable. Some of the writers of our generation will later be looked upon as great, but the vast majority of modern writing will be forgotten, as is true of any era. If you only read modern writing, the quality of what you read will on average be lower than if you only read books that have had time to prove themselves.

Before you can be great, you have to be good. Up until the 20th century, copywork was the primary form of writing education in schools, until we all got hooked on phonics. The Dreyfus model shows that we learn first by imitation, then simple heuristics or rules of thumb, and eventually we learn to use our intuition. It has to happen in that order: you can’t skip a step. It’s through copying others that we begin to notice patterns. From implementing those patterns, we begin to notice when they don’t apply. Finally once we’ve mastered a topic, we can improvise on-the-fly without even thinking. The thing is, we’re all beginners at something. Even celebrated experts in a field don’t have the time to go deep into every topic in their domain, and if they have, it won’t be long until their knowledge is out of date, because everything’s always changing. You swipe file helps you keep on top of it.

Swipe files should be built continuously, whenever you see something worth remembering. However they can be built systematically to aid a specific cause. For example examining all the ads in the Facebook Ad Library for a specific industry, comparing all the homepages of your competitors, or reviewing what differences there are between websites who ranked on the first page of Google for your target keywords. Do the work. Collecting hundreds of examples in a specific domain is guaranteed to break creative block. The human brain is a meme-recognition machine. You won’t be able to help but notice what memes are common or rare. Patterns will start to emerge, and you’ll start to see opportunities to remix.

When you collect ideas for your swipe file, don’t just blindly file them away: attempt to categorize, analyze and reverse-engineer what makes it work. Creativity doesn’t come from the gods: deliberate practice is the best muse. You won’t usually have data on whether a campaign performed, but that’s OK. Seeing an ad run for a long time is a good enough sign it’s working. If the campaigns are being maintained by professionals, they would have turned it off if it didn’t perform. Besides your analysis only needs to be directionally accurate. Your swipe file is for forming hypotheses only: you still have to test the ideas it inspires.

About the Book 📕

← Back to Essays 📝