This post was originally written in Italian by Paolo Gambardella. It has been translated by an AI agent and may contain inaccuracies.
The entertainment industry uses time spent as the primary measure of a product’s or service’s value.
Value = Time × (Revenue / Unit of time)
I’ve been using LLMs and machine learning for a few months now, and in my view this wave of innovation will bring three things: more supply, less pricing power, and above all an audience that compresses everything it can into less time. If that’s true, the formula above starts to crack. Time becomes the scarcest resource, and anyone trying to grow purely by stretching sessions ends up competing with everything else for the same minutes. At that point, what matters is how much each minute a player chooses to give you is actually worth.
Judged by D1 and D7
I’ve worked in mobile games for years, and my job as a designer has always been measured against Retention and Monetisation. The first obviously leads to the second: D1, D7, DAU/MAU. It makes sense — an F2P game survives on the volume of people who show up every day and, crucially, come back. The more someone returns, the more likely they are to start spending serious money. So every feature I designed had to move one of those numbers, or it got cut.
The problem is that some features do their job quietly. A system that makes it clearer what to do next, a small narrative moment that gives a session some meaning, an interface that removes friction — none of these things will shift D7 on their own in an A/B test. So they end up at the bottom of the pile, behind yet another limited-time event that does move a number, immediately. I’ve watched entire roadmaps bend towards whatever was easy to attribute, and the games that came out of that process all had the same face.
The most painful example happened to me last year. I was working with a large client and one of the best teams I’ve ever come across — creative, capable people who genuinely believed in the project. The numbers coming in were interesting, at least from where I was standing. On top of that, we essentially invented a new genre from scratch. No metric was saying the game was doing badly; it was doing well, just not as well as the parent company expected — they were chasing D1 and D7 targets that were frankly unrealistic. The pressure became so intense that, from what I heard later, the project was shelved. A real shame. The way I see it, metrics are there to tell you where not to go — and when they become the only yardstick for judging an idea, they end up killing the good ones too. That company probably had a great product on its hands, and all it would have taken was letting the people get on with it.
A medium you can’t summarise
Video games broadly — not just F2P — are in an interesting position. Think about a film or a podcast: in theory you can swap three hours of film for a two-minute recap. With a roguelike, though, it’s hard to summarise a run… the value is in actually playing it. As a business, that puts us in a privileged spot, and I think it gives us a real edge over other media.
Right now the market is in a consolidation phase, so it keeps repeating formulas it considers proven. That happens in every mature market. And on top of that, as I was saying, new technology in theory lets every person on a team produce more.
Who comes back
I think players have always looked for games that feel unique, and when everything starts to look the same, that search gets more urgent. I wouldn’t be surprised if we started measuring retention differently — looking not just at how many people come back, but which people come back. WHO returns: that’s the real question.
And to understand who comes back, you have to look at what they do when they get there. The player who spends twenty minutes fine-tuning their build, the one who reads every line of dialogue, the one who skips an easy reward because something else interests them more — each of them is telling you something about what matters to them. Optimising, collecting, discovering, making sense of what they’re doing. Build the game around those values and that player stays for years. Treat them as a data point in your D1 average and you lose them without even noticing.
Some will say the industry already looks at “who”: whales, high spenders. For years we’ve optimised for spending depth — how far into their wallet someone was willing to reach. Attention depth is a different thing: how far into the game someone is willing to go, understand it, talk about it, stick with it. The two often overlap, but when they don’t, it’s the second one that lasts longer.
Very often the passionate player — the one who spends, who brings friends along, who stays with you for years — doesn’t stand out clearly on a dashboard showing D1 retention trends. And yet that player weighs heavily on the game’s profitability.
The bet on depth
This is where I think the next shift will come, and soon. A studio that builds its strategy around attention depth might have lower DAU numbers and still win on margins. At the micro level the metric can look a bit ugly, but at the macro level the picture flips.
In practice, that means looking at different numbers: 90- or 365-day retention for your most engaged cohort rather than the D1 average, how many players reach the deeper systems, how many bring a friend, how many create something around the game. These are slower, noisier metrics — and that’s partly why nobody puts them at the top of the dashboard today.
