May 22, 2026 · 5 min read · GameMantra Team

Wallet balance is the churn signal you already collect

Session-based churn signals tell you a player is leaving. Their currency balance often tells you why, and it moves earlier than session count does.

Most work on spotting players who are about to leave runs on engagement data. Sessions per week, days since last login, time in app. These signals work, in the sense that they identify players who are drifting away. What they do not do is tell you anything about the reason, which means the intervention they trigger is usually generic — a push notification, a return bonus, a re-engagement offer that would have been sent to anyone.

There is a second signal sitting in the same database that most studios collect and few read for this purpose. What a player is holding in their in-game wallet, and how that balance has been moving, describes the shape of their experience in a way session counts do not.

What a balance trajectory actually shows

A player's currency balance is the running total of every reward they have earned minus everything they have spent. Read as a series rather than a snapshot, it describes whether the game has been giving them things to do with what they earn.

Three trajectories show up repeatedly, and they mean different things.

A balance climbing steadily with no meaningful spending says the player is earning currency they cannot find a use for. They are still playing — the sessions look fine — but nothing in the game is currently worth what they have. This is a player who is running out of reasons to log in, and the balance shows it before the session count does. By the time engagement drops, they have often been quietly disengaged for a week or more.

A balance sitting near zero for an extended stretch says the opposite. This player wants things and cannot afford them. Every session ends with them short. That is a very different failure — the game has plenty for them to want, and the earn rate cannot keep up. These players churn too, but pushing more content at them makes it worse rather than better.

A balance that oscillates — climbing, then dropping when they buy something, then climbing again — is the healthy pattern. The player earns, sees something worth having, and spends. That loop is the thing an economy is supposed to produce.

The useful part is that all three of these are visible while the player is still active. A rising-with-no-spend balance is an early warning that engagement metrics will not surface for days.

Why this reads differently from a spend signal

It is tempting to collapse this into "are they spending", which studios already track. The distinction matters.

Spend tracks whether transactions are happening. Balance trajectory tracks whether the player has the capacity and the motivation to transact, which are separate conditions. A player with a large balance and no purchases has capacity and no motivation. A player with an empty wallet and no purchases has motivation and no capacity. Both show as "no spend" in a purchase report, and they need opposite responses.

The same collapse happens with paying players. A player who buys currency regularly but whose balance keeps climbing is buying out of habit rather than need — they are accumulating faster than the game gives them reasons to spend. That is a fragile position. When they notice the pile, the habit usually stops. A player who buys currency and immediately spends it is buying because they want something specific, which is a much more durable pattern.

Reading balance alongside spend separates those two paying populations, which a revenue chart shows as identical.

The comparison that makes the number legible

An absolute balance figure means nothing on its own. Ten thousand coins is a fortune in one game and pocket change in another, and it means different things at day 3 and day 300 of the same game.

What makes it readable is comparing a player's balance to the cost of the things currently available to them. A balance that sits well above the most expensive thing they can buy is a player with nothing left to want. A balance that sits well below the cheapest meaningful upgrade is a player who is stuck. The ratio between what they hold and what the game is asking for is the number worth watching, not the balance itself.

This ratio also travels across your catalog in a way a raw balance does not. A player in one game holding four times the price of the next upgrade and a player in another game in the same position are in comparable situations, even if the underlying numbers differ by orders of magnitude.

The same comparison applied over player tenure tends to reveal design problems that were invisible in aggregate. If the ratio drifts upward across every cohort past a certain point in progression, the game has a content ceiling there — players reach it, keep earning, and stop finding things to buy.

See how we read economy signals across a live game →

What each pattern asks you to do

The reason to separate these signals is that they call for different responses, and sending the wrong one costs you the player faster.

For the rising-balance player, the answer is something to want. A new tier of content, an item that costs meaningfully more than anything currently on offer, a reason for the pile to have a purpose. Sending this player a discount is close to insulting — price was never their obstacle. Sending them more currency actively worsens the problem.

For the empty-wallet player, the answer is either a better earn rate or a cheaper entry point. Something they can actually reach. This is the player a well-placed small offer serves genuinely, because their constraint is real and the offer removes it. Sending this player new expensive content deepens the frustration.

For the healthy oscillating player, the answer is usually to leave them alone. The loop is working. Interrupting a functioning economy loop with an intervention designed for a struggling player is a way to create a problem that did not exist.

The wider point is that "this player might churn" is not actionable on its own. Two players with identical session decline can be leaving for opposite reasons, and the wallet is often the cheapest place to tell them apart — because you are already recording every transaction that produced it.

Talk to us about reading your economy's signals →

Share this post

See what this looks like for your game.

SDK for Unity and Unreal. A 20-minute call to walk you through it.

Book a demo