Why some apps get a 70% “yes” and others get 25%, asking the exact same question

Apple’s App Tracking Transparency (ATT) is one screen. Two buttons. “Allow” or “Ask App Not to Track.” Apple wrote it, every app shows it word for word, no one gets to edit it.

And yet some apps get a 25% opt-in rate. Others get 70%.

Same screen. Same two buttons. Same user.

So the screen isn’t the variable. Something before the screen is.

I’ve been chasing this question since I first started writing about permission microcopy push notifications, location access, cookie banners, all the same fight for a “yes.” ATT is just the cleanest version of it, because Apple controls the ask so tightly that there’s nowhere left to hide. The apps that win aren’t writing cleverer copy inside the dialog. They’re doing something before it ever appears.

The ATT problem is a permission problem

ATT is structurally identical to every other permission prompt I’ve written about: push notifications, location access, camera access, cookie consent. The user is being asked to give up something (tracking transparency) in exchange for something (a better experience, in theory).

The system prompt is fixed. Apple controls it. You can’t change the words “Allow App to Track Your Activity” or “Ask App Not to Track.” But you can control everything that happens before the system prompt appears.

And “before” is where the entire game is played.

The apps with high opt-in rates all do the same thing: they show the user what they’ll get, and they do it at a moment when the value is self-evident. The apps with low opt-in rates do one of two things: they show the system prompt immediately (no pre-prompt), or they show a pre-prompt that explains tracking in technical terms without connecting it to user value.

Here’s the pattern, in three steps.

Step 1 — don’t ask on day one. Wait until the person has felt personalization work. Browsed a few products and seen a recommendation land. Read some articles and watched the feed sharpen. Played a level and gotten a reward that actually fit. Now, when you ask “can we keep doing that?” — they know exactly what you mean. They just lived it.

Step 2 — before Apple’s dialog shows up, show your own screen first. Plain words, your framing, no jargon. Something like:

Notice what that does. States the benefit. Names the mechanism honestly. And makes “no” feel completely fine to choose, not a trapdoor.

Step 3 — one button on your screen: “Continue.” Tap it, Apple’s dialog fires immediately. Whatever trust you just built carries straight into that “Allow.”

Run it as an A/B test — pre-prompt versus dropping the Apple dialog cold — and the gap holds at 2 to 3x. Not a nice-to-have lift. A completely different outcome from the same question.

Consent becomes meaningful the moment you’ve shown value before asking for access.

Why this isn’t just an Apple problem

If you build for ecommerce or martech, this number is quietly load-bearing. No tracking consent means no ad attribution — a merchant runs Facebook ads and can’t tell which one paid off, so ROAS goes dark, and they blame the tool. Push notification targeting shrinks because “browsed product X” stops existing as an audience. Cart-abandonment email — “you left this in your cart” — needs the exact cross-device thread that consent provides. Cut the thread, the merchant loses the one thing they paid you for.

So the opt-in rate isn’t an ad-tech vanity metric sitting in someone’s dashboard. It’s the ceiling on what your product can actually deliver.

The same pattern, off of ATT entirely

Push notifications on a shopping site — don’t fire the browser prompt on first visit. Wait for an add-to-cart. Then: “Want a heads-up when this goes on sale?” Now there’s a specific item and a specific reason to say yes.

Email popups — skip the page-load trigger. Wait until someone’s browsed two products or spent 30 seconds on site. By then, “items like these” means something to them.

SMS at checkout — split the two asks. Order updates first, high value, zero resistance. Marketing consent after the order confirms, framed around what they’ll actually get.

Show value. Explain the exchange. Respect whatever they decide. Same three moves, every time.

One thing worth being honest about

Opt-in rates also shift by market, and that part isn’t a UX fix. India and Southeast Asia are more used to the free-with-ads trade, so a clear value story lands well. Europe carries eight years of GDPR-shaped caution, and opt-in runs lower across every consent type, not just this one. The US sits in the middle — wanting the personalization and resenting the tracking, at the same time.

Which means localizing the pre-prompt isn’t translation. It’s re-making the case for what that market already believes about the trade.

The actual principle underneath all of this

ATT was never really a policy problem. It’s a permission design problem, and permission design comes down to one question:

Did you earn the right to ask?

You earn it by showing value first — not by explaining your tracking stack, not by apologizing for needing the permission. By showing someone what they get, in words they actually care about, right when it’s obvious.

The 70% apps aren’t manipulating anyone. They’re just asking at the right time.

Thank you for reading and keep creating.

Yours truly, Riya

Why some apps get a 70% “yes” and others get 25%, asking the exact same question was originally published in Bootcamp on Medium, where people are continuing the conversation by highlighting and responding to this story.