Understanding residual cohort value: The hidden asset in mobile gaming and apps

• Residual cohort value represents the future revenue already sitting inside a mobile game’s existing players.
• Most finance teams do not explicitly track or report this value in their standard financial reporting.
• Longer retention and monetisation tails significantly increase the residual value embedded in a game.

Martin Macmillan is the founder of the Mobile Finance Collective and this article was first published as part of The Capital Stack newsletter.

There is a number your finance team almost certainly isn’t tracking closely enough. It doesn’t appear on your balance sheet. It doesn’t show up in your P&L. And yet it may represent the single most significant store of value in your entire mobile business.

It’s the residual value locked inside your existing user cohorts.

Understanding it could change how you think about your business, how you present it to your board, and ultimately, what it’s worth

The flywheel you’ve already wound up

Think of mobile user acquisition as a giant flywheel. Every dollar you spend on UA puts kinetic energy into that wheel. The longer you spend for, and the more you spend, the more momentum it accumulates.

When you stop adding energy, whether by choice, budget constraint, or simply as a thought experiment, the wheel doesn’t stop immediately. It spins out. Gradually. Over months, or even sometimes years.

What determines how much energy (revenue) comes out the other side? This is largely down to two key factors:

1. How long do your players or users stay engaged in your game or app

2. Their propensity to keep spending over that period

This is where genre and product type matter enormously. In mobile gaming, at one end of the spectrum, a hypercasual game would likely exhaust its cohort’s meaningful revenue potential within thirty days.

Gameplay is simple, content is thin, and the commercial model is built around volume of ads served and velocity rather than depth. At the other end, you have subscription apps that become genuinely habit-forming, such as productivity tools, health apps, or casual games where users have been paying monthly for years. Match-3 titles, social casino games, and certain RPG genres sit in a similar place. These are products with a long flywheel tail, or long cohort duration.

The longer the tail, the more energy is still in the system and the more value is sitting there, unrealised and unmeasured, in the cohorts you’ve already acquired – those which you have already paid for and have already partially monetised.

Why this matters: Two unavoidable moments of truth

There are two moments when this residual value stops being theoretical and becomes very real.

The first is when you sit down with certain types of lenders. Outside of cohort financing, where lenders use your previous cohort data and provide capital to acquire new cohorts expecting them to follow the same pattern, other types of credit providers who operate in the gaming and apps space don’t just look at your current revenue run-rate or your cash position.

They want to understand the expected cash generation of the business as it already exists, even before you spend another dollar on UA. The residual value in your cohorts is effectively collateral. It represents a stream of future cash flows that can, under the right conditions, be borrowed against. Banks typically cannot price this type of collateral so it falls to more sophisticated private credit funds with better vertical knowledge of mobile.

The second is when an acquirer comes calling. The most sophisticated buyers such as strategic acquirers, specialist PE firms, and roll-up vehicles that understand the space don’t simply apply an EBITDA multiple based on prevailing market conditions and comparables to price an acquisition.

They look to build a more sophisticated model, a part of which is wanting to understand the terminal value of your existing user base: how long those cohorts will keep monetising, at what rates, and what the present value of that stream looks like at various discount rates.

A CFO who can hand them that analysis which is pre-built and clearly presented, is speaking the same language. A CFO who can’t could be leaving meaningful value unrecognised on the table.

Step 1: Monetisation profile of a single cohort

The methodology starts simply: pick a representative cohort and map what it actually produces over time.

A critical point before we go further. This analysis is explicitly not around ROAS. 

The advertising spend that acquired these users is already sunk so it has no bearing on the value of what those users will generate from this point forward. What we’re measuring is purely the revenue the cohort produces over its lifetime. The ad cost is irrelevant to the asset for the purposes of this analysis.

For an IAP-heavy mid-core game, the daily revenue profile of a cohort of users looks something like this. There’s a pronounced early spike driven by D1 and D7 spenders – potential whales and early converters who monetise quickly – followed by a long, gradual decay as the cohort matures. Some users churn; the remainder monetise at decreasing but persistent rates over a multi-year tail.

The solid line is observed data, in this case 270 days of actual cohort monetisation. The dotted line is the projection, extending out to D720 using a curve-fitting methodology applied to the observed history.

For the projection itself, several methodologies are available, such as least squares regression, power law decay models or parametric LTV models built within your MMP. The right choice depends on the volume and quality of your cohort data.

The more historical data you have, the more confidence you can place in the projection. As a practical note, apply a minimum seasoning period of 30 to 60 days before incorporating very new cohorts into the model. Very early-stage cohorts lack sufficient shape to forecast reliably.

One final observation on the shape of this curve: notice how the incremental value added between D360 and D720 is materially smaller than the value accumulated in the first D180. The curve flattens, which has important implications for how you weight different time horizons when presenting residual value and for how a lender or acquirer will discount the longer-dated tail given a changing risk profile.

Step 2: Mapping the cohorts

A single cohort tells you about one month’s UA spend. But for this approach to be applicable, studios need to have been acquiring users continuously over some period of time at a reasonable level of spend, likely a minimum of 9-12 months for the flywheel to contain a meaningful level of residuals.

The next step is to lay all of those cohorts out on a common timeline, each running its own decay curve from its own acquisition date. Earlier cohorts have more observed history; more recent ones are predominantly projected. The picture looks like this. For simplicity a constant level of spend and UA performance has been assumed.

These are the individual components of the flywheel broken out and made visible. Each line represents a monthly cohort, running its own revenue trajectory from its own acquisition date. The earlier cohorts have mostly played out their steep early value and are now generating from their long tails. The more recent cohorts still have much of their monetisation ahead of them.

Step 3: The full picture – Rebasing and stacking

In order to consolidate the picture, the cohorts need to be rebased to today and then stacked.

Add every cohort’s forward projection together and you get the aggregate daily revenue profile of your existing user base – the total expected output of the flywheel from today forward, assuming zero new UA spend.

The area under the curve is the residual cohort value. The revenue milestones below the chart show the expected future value of the cohorts at 180-day intervals out to 720 days. A good CFO should look to discount back these (or realistically at a more granular level) using an assumed cost of capital in order to get a NPV (Net Present Value) of the residuals at each different point.

Of course, the cohorts may reach out way beyond 720 days in certain genres, but of course with greater extended time periods of forecast comes greater uncertainty.

These are the numbers that don’t appear anywhere in your standard financial reporting, but probably should.

Putting it in front of your board

This analysis deserves a standing slide in your monthly board pack. It should not be buried in an appendix, but presented as a core metric alongside revenue and cash position. Updated monthly, it does several things at once. It reframes the UA as a conversation. Rather than a pure cost line, accumulated UA becomes a balance sheet-adjacent asset with measurable, model-backed value.

It gives your board a forward-looking view of high conviction cash generation that most P&L’s can’t provide – what the business will produce even if you stopped spending today (providing sufficient OPEX remains in the company to keep the lights on). And it positions your finance function towards thinking about the business in exactly the way that more sophisticated external parties will think about it when it matters most.

Residual cohort value is not a future aspiration, but a present reality, sitting inside your existing user base, in many cases substantially larger than most finance teams appreciate. Measure it. 

Track it. Present it to your board. And begin to think about what it might mean for the capital available to you to fund not only additional UA spend, but also studio acquisitions or allocating capital to internal projects through cross-collateralising the residual value across all your titles.

The CFOs who build this capability early will be significantly better prepared for the conversations that define what their business is actually worth.