When to Stop Researching One Advertiser

I used to keep a tab open for every advertiser I found interesting. I’d check their landing pages, look at their angles, scroll their feeds, and tell myself I was being thorough. Most of the time I was just avoiding the part where I had to decide.

Now I give myself a stopping rule before I start. The rule is simple: I look until I can answer three questions, and then I close the tab.

The first question is what they’re selling and who they think they’re talking to. The second is which creative angles keep showing up. The third is what a reasonable next test would be for my own account. Once I can say all three out loud, more scrolling rarely changes the answer. It mostly adds detail I won’t use.

One way to judge whether you have enough is repetition. If the same offer and the same angle keep appearing across different creatives, you probably have the pattern. If every new ad tells you something completely new, that’s a sign the advertiser is testing hard and you’re still early. There’s nothing wrong with that, but it’s worth knowing which situation you’re in.

Diminishing returns look predictable. The ads start to blur, you’re saving screenshots you’ll never open again, and you can’t remember why you saved the last one. That’s the signal. I stop there.

What I do instead of continuing is write a short note. Mine has four lines:

• Offer and audience, in a sentence.
• The angle I think is carrying the account.
• One thing I’d test against my own current creative.
• A date, so I know when I looked.

That note is worth more than fifty more screenshots because it’s already in my own words. If I use a native ad spy tool to pull examples, I keep the note right next to the few ads that actually taught me something, not the whole pile.

The next test matters more than the research. I pick one variable, change only that, and let it run long enough to mean something. Say I test a new opening line against my current one. I don’t touch the image, the offer, or the audience. Small, clean results are easier to trust.

The last thing I’ll add: set a timer. Ten or fifteen minutes is usually plenty for one advertiser. If I hit the timer before I can answer the three questions, I move on and come back later. Most advertisers don’t need to be fully understood. They need to be useful for about ten minutes and then left alone.

— Ryan Brooks