Cross-title player value across a studio portfolio

Studios with more than one game usually measure each one separately. Each title has its own retention, its own revenue, its own churn.

That is the right way to run a game and the wrong way to run a portfolio, because it makes every player movement between your own titles look like a loss on one side and unattributed acquisition on the other.

What per-title measurement hides

If a player stops playing your first game and starts playing your second, three things happen in the reporting.

The first game records a churn. The second records an install, usually attributed to whatever marketing channel happened to touch them, or to nothing at all.

Nobody records that this was the same person and that the studio kept them. From the portfolio's perspective nothing was lost, and from the reporting's perspective a player left and a stranger arrived.

The consequences show up in decisions. The first game is judged as leaking players and gets investment in retention, some of which will be spent trying to hold players who would be worth more in the other title. The second game's acquisition looks better than it is, because part of it is internal movement being counted as new.

The measurement that fixes it

The requirement is being able to recognise the same person across titles, which means a shared account or identity layer. Studios that added accounts per game find this hard to retrofit and it is worth doing anyway.

With that in place, the useful figure is portfolio-level: what a player is worth across everything you operate, over their whole relationship with the studio. That number is higher than any individual title's figure for the players who move, and those players are frequently your best ones — engaged enough to try your next thing.

The second useful figure is the retention rate at the portfolio level. A player who leaves one title for another has been retained by the studio, and knowing what share of your leavers do that changes how you read every individual game's churn.

What it changes about decisions

The first change is in how you treat a game in decline. A title losing players is a problem if those players leave the studio, and an opportunity if they move to another of your games. Those call for opposite responses — fix the leak, or make the path to the next title as smooth as possible.

Studios without portfolio measurement default to fighting every churn, which sometimes means holding a player in a game they have finished rather than moving them to one they would enjoy.

The second is in acquisition. If a share of one title's installs are internal, its true external acquisition cost is higher than reported and its channel performance is being flattered. Deciding where to spend based on that is deciding on a number that includes your own players arriving for free.

The third is in what gets built. If a particular kind of player moves reliably between two of your titles, that is a real product finding about what those players want, and it is invisible per-title.

See how we look at players across a studio's games →

The part that requires restraint

Cross-title visibility creates an obvious temptation: actively pushing players from one game to another.

Done carefully this is one of the better acquisition channels available, because the player already knows the studio and the recommendation has context. Done carelessly it damages both titles — the first because players feel pushed out of a game they were enjoying, and the second because they arrive resentful.

The distinction is whether the move is offered at a moment when the player is actually finished. A player who has completed the content, whose engagement has been declining for weeks, is a reasonable candidate. A player in the middle of a good run is not, and interrupting them to advertise another game costs more than it produces.

The other restraint is on data. Recognising the same player across titles for measurement is different from combining everything you know about them into a single profile used for targeting, and the second has privacy obligations that vary by market and by the ages involved. The measurement value comes almost entirely from knowing that two accounts are one person, which is a much narrower requirement than knowing everything about them in one place.

Getting the measurement right and being conservative about the targeting captures most of the value with a fraction of the exposure.

The other measurement that becomes available is which titles feed which. Over time, movement between your games forms a pattern — players who play one tend to move to a second and rarely to a third. Knowing that shape tells you which titles are genuinely complementary and which compete for the same person, and it is directly relevant to what you greenlight next. A studio building a third game that draws from the same audience as its first two is splitting an audience rather than growing one, and the movement data is where that shows up before the revenue does.

Talk to us about portfolio-level measurement →