Going the distance
to build games
people love.

When the traffic model broke I had two doors. One: lay off most of the team and pivot to burst-session mechanics we had no experience in and no appetite for. Two: reuse the engine, reskin Rome to Sengoku Japan, move to a leaner model, and move fast. We greenlit Daisho on 23 January 2023 and shipped it that June — concept to live in five months. It kept the team whole. Whether it beat handing the rest of the money back to investors, I still can’t tell you.

Early 2023, Google Play swapped most of its organic auto-collections for ad inventory. To a player the home page looked identical; to us, free traffic had silently become paid traffic. A low-LTV player from Vietnam covers his keep against staff and servers — bolt a CPI onto his head and he’s a hole in the P&L.

The launch went well enough that I could have raised again at a $60M valuation. I passed — we had cash in the bank, inflation was high, and I wanted to build, not dilute. Months later the ad market turned and the number would never come back. I’d make the same call and I’d lose sleep over it again.

My first real touch with mobile was a MOBA that couldn’t pay for itself. Marketing budgets were pulled, so I had to learn organic growth or watch it die. That necessity became the method: one proven survival-RPG template, reskinned into settings the big spenders find too small to bother with — Rome, then Sengoku Japan, then the Viking north.

In 2019 I wrote that Google Play quietly mines user reviews for keywords — including the competitor names, abbreviations and localized titles players type but a developer is barred from using. Sentiment everyone treated as noise was already feeding search and browse traffic, for free. Almost nobody read it that way. Years later that one observation is what a whole company runs on. The edge wasn’t a tool nobody had — it was reading something everyone already had.

Forty-six updates to Gladiators in fifteen months, not one rejected for its store page. The forty-seventh drew a reviewer who invoked the vague “screenshots must reflect gameplay” rule and, with it, overruled every colleague who had approved the page before him. The biggest titles ship pure promo art with no interface at all and get featured for it. I don’t dispute Apple’s right to write the rules — only a rule read one way for them and another for whoever has the least standing to argue it.

The order of operations flipped while I wasn’t looking. We used to build the game, then work out how to sell it. Now the selling decides the building: micro-moments instead of marathon sessions, because long sessions quietly cost you retention; creative proven with a few AI stills and a tiny test budget before a line of feature code; everything measured against what a player is worth on day thirty, not day one. The ad network learns on your creative — so design things it can learn on.

A well-followed voice posted that Apple reads and indexes the text inside your screenshots. It fit nothing I knew — its reviewers reject screenshots for having too much text, and its search matches exact terms; it doesn’t read pictures. So I went to the comments for the few people who’d actually know, expecting a correction. I found applause, and quotes that could have been about anything. The room of people with millions of organic installs behind them is small. The room reposting them is not.

In 2019 I wrote that the coming privacy wall would quietly blind paid user acquisition — Apple killing third-party tracking, attribution data drying up, and the targeting edge that made campaigns predictable going with it. That was two years before ATT actually landed. The conclusion wasn’t that advertising was dead. It was that the advantage would move to whatever privacy couldn’t take away: store-page visibility, real relationships with the platform’s editors, and the data you own because it lives inside your own game. I’ve built every studio since on the parts nobody can revoke.

I once took apart a Berlin studio’s collapse — north of eight million euros, four years, two decent games, gone. You didn’t need the gossip to see it coming; you needed the arithmetic. A team that size burns a fixed number every month. Divide it by what a casual player is really worth after the platform’s cut and the taxman, and you get the revenue you have to clear just to survive — a number their games were never going to reach. The hard part isn’t the math. It’s running it before you’ve fallen in love with the idea.

There’s a sin I keep meeting in founders who were players first: the proud vow that the game is never pay-to-win, only skins, the most player-friendly studio that ever shipped. Staying off the predatory stuff is right. But vanity items top out near half a percent pay rate, each one sold exactly once, and they quietly tax the people who built the game. The unromantic truth is that the players who pay are usually your most engaged — paying cohorts retain better than the average install, not worse. Selling real value to people who want it isn’t the opposite of caring about players. It’s how the team that made the game gets to make another.

A recruiter once asked me how to spot a real ASO hire. The tempting move is to weigh their biggest past numbers — installs, a famous launch. But hand a mediocre marketer a war chest and they look brilliant; starve a brilliant one of budget and they look ordinary. So I stopped asking what they shipped and started asking how they’d build a keyword core, why the App Store and Google Play reward opposite things, what actually belongs in the first screenshot. Those answers don’t depend on whose money they were near. The trophy numbers do.

I do a lot of teardowns on games that underperform, and the same wound shows up again and again: a cold open of skippable cutscenes and explanation windows, icons that mean nothing yet, a player tapping through a tutorial for systems they can’t use — churning before the first real moment of play. Fun first. The map, the shop, the crafting tree can all wait until the player has felt why the game is worth understanding. You teach the rules of a game people already want to keep playing, not the other way round.

On Planet of Heroes — a real-time mobile MOBA, an Apple Trend Game of 2017 — we shipped a story mode from day one. It felt generous. What it actually did was split the player base clean in half: a large share never touched the PvP the whole game was built around, which meant longer matchmaking, emptier brackets, a worse game for everyone who came for the fight. Even League of Legends shows it in its queue times. If your game needs a crowd in one place, think hard before you build a second place to stand.

A game that’s pure, uninterrupted fun is a quiet killer of revenue — the player never stops long enough to buy anything. So you interrupt: a wait, a wall, a prompt. But interrupt someone who’s enjoying themselves and the irritation usually sends them out of the game, not into the shop. And on the rare occasion they do pay to skip the wait, you’ve just spent the very time you had left to sell to them again. Fun and monetization are pulling on the same minutes. I learned it most sharply in real-time PvP, where the core loop is the most fun and the least interruptible — but it holds for almost any game. The whole craft lives in the seam between too little friction and too much.

The Google Play team came through and made our flagship their showcase — install it, look, this is what fully cross-play 4K looks like without three hundred engineers behind it. Five stars, real admiration. But all of it was in service of Google Play Games for PC: a product their own company barely resourced, never made performant, and plainly didn’t believe could dent the PC market. So we were the proudest exhibit in a hall nobody upstairs meant to keep open. The people on the ground loved the game. The institution paying them had already moved on. Applause from the field is not commitment from the firm — and being the best thing on a neglected platform is still being on a neglected platform.

I think of myself as a marketer before anything else, and to me marketing isn’t decoration bolted onto a finished game. It’s translation. The desire is already sitting in the player’s head — for a world, a fantasy, a feeling they want more of. The whole job is to render that desire faithfully into the picture they scroll past, the video they watch, and finally the game they open. Get the translation right and growth stops being something you buy and becomes something you earned by listening. Build the game the desire was already asking for, and the marketing was half-done before you started.