Your Tall Pin Starts the Race With a Handicap

Pinterest just told us in plain words that they stopped showing your best ad to everyone. And if you've been making your pins taller to get noticed, they said something in one line that's for you too.

Here's the idea. Pinterest no longer picks your best ad. It picks a different version for each person who sees it, and it docks any version that hogs the screen.

The Thing We All Believed

I tested five pins, one won, I'll run that winner. I've said it myself, and for years that's how Pinterest worked too. The person ready to buy at 2pm and the person scrolling in bed at 11pm saw the same exact pin. Same slice of cake for every kid at the party, whether they wanted chocolate or not.

Pinterest Stopped Showing Your Best Ad to Everyone: Spotify

Pinterest Stopped Showing Your Best Ad to Everyone: Apple

Meet the Helper

There's a white paper from Pinterest's own ads team. They call the system Pin DCO. I'm going to call it the helper, because that's what it is.

Picture a helper standing at the front door of a store. You hand them every version of your ad. Every time someone walks in, the helper reads who they are and picks the one version that person is most likely to click. Every person, every time, in a fraction of a second.

Why Your Tall Pin Starts Behind

The feed is columns, like shelves in a store. A tall pin takes more shelf space and shoves three other pins off the screen until someone scrolls.

Pinterest now counts what those three pins lost, so the helper gives every taller version a penalty. It can still win. It has to earn that space by performing a lot better than the shorter version would.

They tested it. Clicks ran 9.5% with the penalty on and 9.8% with it off, basically the same, while ads only got 5% taller instead of 7%. Meanwhile the older systems, the ones chasing clicks alone, got a 1.5 to 1.7% lift and the rest of the page went down.

If you're spending money on ads and your creative is one tall pin from a template shop, you're paying to run a version that starts the race with a handicap every single time.

What the Helper Actually Wants

Choices. That's the whole answer. One image and one title means the helper has nothing to pick from. Aim for three to five images, two to three titles, and both layouts if your offer fits both.

Now the nuance: the versions have to be genuinely different. The paper spells out what gets graded. Image size. How big the product sits in the frame. How much it pops off the background. How tall the pin is.

Same photo with a new font is the same pin. A close-up versus a lifestyle shot is a different version. A white background versus that product sitting on a kitchen counter is a different pin.

Pinterest even showed us which layout went to whom. The two-part layout, one hero image with smaller product shots underneath, went mostly to high shopping intent. The single mood-and-taste image went to browsers. Architecture, design, services, people still dreaming.

So build one shopping version and one browsing version, and you stop paying to guess who's coming in the door.

Five to Seven, Not Forty

A small slice of the time the helper shows a version that isn't the predicted winner, just to see. That's how brand new pins get real data, The engineers are clear that gains flatten after the first handful. So 40 versions on day one is mostly paying for storage.

One more line, and I think it's the real reason they built this. Pinterest says outright that AI image generators are why. They expect ad versions per campaign to explode, and they wanted a system that could sort hundreds without the app slowing down.

You don't need hundreds. You need five or seven that are different.

We're seeing people get lazy with creative, using AI only, skipping actual images of the products and the things you created, cooked, made, slaved over. In my agency every pin design is made by a human. We use AI for research and to speed up putting words together, and it's never the final output.

Final Pin Drop

This paper covers ads only. The regular feed runs on a different system. I'll always tell you when something is a guess, so here's mine. Same company, same shelf, same scoreboard, which is whether a person saved or clicked during their whole visit. Standard 2:3 stays your home base and a taller image has to earn it organically too. Make several genuinely different pins per page, and make the main thing in the picture big and obvious.

Go open your ad campaign and count the versions. Not the pins in your account, the versions inside that ad.

If the answer is one, congratulations. You just found your afternoon.

📌 Book a PinChat → laurarike.com/pinchat