The live-service correction and what smaller studios learn

The live-service investment thesis that drove publisher strategy from roughly 2018 through 2023 has corrected sharply. After high-profile failures — Sony's Concord cancellation, Warner Bros. Games' MultiVersus reversal — major publishers now typically carry one or two genuine live-service productions in active service rather than a slate, and publishing plans have shifted back toward finite, single-purchase titles. (Gameindustry.com)

It would be easy to read this as the live-service model failing, and easier still to read it as permission to stop worrying about ongoing operation. Both readings are wrong, but the second one is expensive.

What actually got corrected

The thesis that failed was not "games benefit from continued operation". It was "a publisher can run many simultaneous live-service titles, each requiring a permanent team and a permanent audience, and the audience will be there for all of them".

That second claim was always doing the heavy lifting, and it was a claim about attention rather than about game design. A live-service game asks for a recurring commitment from its players. Players have a limited number of those to give — the number of games someone actively keeps up with is small and does not expand because more games want to be on the list.

So a slate of live-service titles from one publisher was, in part, those titles competing with each other for the same finite commitment, on top of competing with everyone else's. The correction is a recognition of that arithmetic, not a verdict on whether operating a game is worthwhile.

Meanwhile, in mobile, the concentration went the other way: the large majority of in-app purchase revenue now flows through titles under active operation. (PocketGamer.biz) Both things are true at once, and they are not in tension. Operating a game works. Operating eight of them at once, each demanding to be someone's main game, does not.

The part that transfers to a small studio

A studio with one or two titles is not exposed to the failure mode that got corrected. It has nothing to cannibalise. What does transfer is the underlying observation about commitment.

If your game is designed to require daily engagement to stay worthwhile — daily reset, daily obligation, progress that decays if you miss — you are asking to be one of the small number of games a player keeps up with. That is a viable position and plenty of games hold it, but it is a competitive slot, and losing it is usually permanent. A player who falls behind on a game built around not falling behind rarely returns.

A game that tolerates absence is competing for a different and much larger slot. Players can come back after two weeks and still find their progress meaningful, and the game can be one of several they rotate through. The revenue per player is generally lower, and the population it can draw from is much larger.

The publisher correction is partly a rediscovery that the first slot is scarce. A studio deciding which one they are building for is making the most consequential design decision they will make, and it usually gets made implicitly through a hundred small pacing choices rather than deliberately.

What does not transfer

The temptation in a correction is to conclude the safe move is the finite, single-purchase product. For a console publisher with a large budget and a marketing apparatus, that is a coherent strategy — they can make a substantial premium game and sell it.

For most small mobile studios it is not the same trade. The economics of a paid mobile title are difficult in ways unrelated to whether live service is fashionable, and the discovery problem does not get easier. "Ship it and move on" was not working for small studios before the correction, and nothing in the correction improved it.

The other thing that does not transfer is the assumption that operation requires the staffing the failed slate implied. Those productions carried large permanent teams because they were producing bespoke content continuously. A game whose ongoing operation comes largely from systems that respond to players rather than from hand-built weekly content has a very different cost structure, and that structure is what makes continued operation available to a team of ten.

See how continuous operation works without a permanent content team →

The question worth taking from it

The useful takeaway is not a direction. It is a question that the failed slate answered badly and that any studio can answer well.

What is this game asking of a player each week, and is it asking for more than the game gives back?

A game that asks for daily attention and returns a genuinely good session each day earns its slot. A game that asks for daily attention out of retention mechanics — streaks, decay, timed windows — while the session itself is unchanged is asking for commitment it has not earned, and players eventually notice.

The publishers that pulled back were, among other things, discovering that several of their titles were in the second category simultaneously. A single studio with a single game can check that directly and much more cheaply: look at what the game demands, look at what a session actually delivers, and be honest about whether the exchange holds up for a player who has other options.

Talk to us about designing around real player commitment →