Hi Reader -

Years ago, I was working on a word game, and we were killing it in the US. So we ran an experiment in the UK. After all, they speak English too, right?

I think it was George Bernard Shaw who said: England and America are two countries divided by a common language. (Whoever it was that said it, I was about to prove it.)

Anywayyy for our word game, the UK did even better than the US. CPI was strong. Cost per first puzzle played was strong. D1 ROAS was extremely strong, and D7 held up too.

There's always someone in the meeting who says, "We're crushing it, let's double the budget." That week, it was me. I also smugly said "there's a reason it's called the Queen's English."

I'd love to tell you I was the skeptic in the room. I was the guy with the budget slide.

Then we looked past day seven for our word game.

People installed, played, paid, and then they just left. Everything after D7 was terrible. My first move was to defend it. No, no, everything's fine. Maybe it's a Facebook glitch. Maybe it's a platform thing. Maybe it bounces back next week.

It did not bounce back next week.

Somebody said "UK users are just different," the way people say it when they want the meeting to end. They were right. We just didn't know how right.

Because I mean, what the hell was going on?

Same game. Same ads. Same language. Every number we had said these were great players. And the great players were disappearing. So we stopped staring at the dashboard and zoomed out.

The ads were fine. The game wasn't built for British players.

We had two problems.

The first was spelling. The puzzles used American spelling. So picture a British player, raised on the Queen's English. They type "colour," the way their teachers drilled it into them. They type "favourite."

And our game tells them they're wrong. Every day. Like a schoolteacher with a red pen and an American accent. A lifetime of spelling, marked wrong by an app they'd downloaded that week.

The second was the references. A lot of the puzzles leaned on American TV shows and American sports. So now they're also being quizzed on The San Diego Padres and the World Series(which, it turns out, isnt quite the world championship of baseball).

And none of it showed up on day one metrics.

On day one, you're just trying the game. You're not reading closely yet. One odd spelling is a quirk.

Then you get into the puzzles, and it starts to feel like a strange new language. The game says your spelling is wrong. The clues are about somebody else's Saturday nights.

Word games work because they make you feel clever. Ours made British players feel foreign(and stupid) in their own language. Nobody keeps paying to feel stupid.

So you leave.

We cut the UK. The budget I wanted to double went to zero. From "double it" to "kill it" in about two weeks. Fixing it meant localization, and localization was a whole new project, not a campaign tweak. We decided to come back to it later.

What I took from it: early numbers tell you the ad found people. They don't tell you the product works for the people it found.

I wrote about the flip side of this once, with a hat in Chile. One cultural detail on an app icon, done right, lifted conversion by double digits.

If you're spending $50k+/month and want someone to read your numbers past day seven, market by market, reply "D7" and we'll talk.

Or don't. I'm not your mom.

Later,
Shamanth

P.S. I'll be speaking at Business of Apps Berlin. If you're there, say hi. I have a couple of complimentary tix, so if you want me to hook you up, hit reply and ask!

P.P.S. Deconstruction of the week: I pulled 2,856 of Suno's ads and broke down what they're doing. The full walkthrough is here.

Here are some goodies from our socials:

​We tested 80 ad variants in 2 months and cut cost per trial by 32%​

​Should you test the audience or the creative first?​

​Before a video ad goes live, I'd score it on these 6 variables​

​How we grew profitable Facebook spend 110%+ for a game studio​

​79% of users check reviews before downloading an app​

​Suno is hiring a creative strategist. Here's the ad I'd make for them​

​The script can make or break the ad​

​AI in 2026 reminds me a lot of mobile UA in 2012​

​Can you make hundreds of photoreal AI skits a week without them turning into slop?​

​The AI creative debate is splitting into 2 camps​

P.P.S. Yes, "colour" in the preview line is on purpose. Please don't reply to correct me. I've been corrected enough.