Whale segmentation beyond the binary: what 5% / 65% means

Mobile gaming's revenue concentration has been an open secret for years. A small share of players produces the majority of revenue. The specific numbers cited in 2026 industry reporting put the concentration even sharper than the historical reads — around 5% of players generating roughly 65% of in-game purchase revenue. In many genres the ratio is steeper still, with 1–2% of players accounting for more than half of the take.

The numbers themselves aren't the news. The news is that the binary view of player monetisation — payer vs non-payer — is no longer granular enough to make decisions with. Studios who segment on the payer/non-payer line treat 5% of their base as a single homogeneous group, when that 5% itself contains the entire interesting structure of how revenue actually accumulates.

The 2026 conversation in the analytics community is about what comes after the binary. The answer that's emerging — sequence-based segmentation, compound behavioral criteria, intent profiles — is more useful than the old labels, and the studios moving toward it are extracting meaningfully more value from their highest-engagement segments.

What "whale" actually is

The word "whale" gets used loosely, and the looseness costs studios money.

In its strict definition, a whale is a player who has spent above a high threshold in the game — often $100 or more in lifetime spend. That definition catches a real population but mixes together very different player behaviors inside it.

Some whales arrived at $100 lifetime spend through one large transaction. They bought a $99.99 bundle once and never spent again. The lifetime spend says whale; the behavior says one-time impulse buyer.

Other whales reached $100 across 50 transactions of $2 each, over several months. The lifetime spend says whale; the behavior says highly engaged frequent purchaser.

Still others reached $1,000 lifetime spend through 50 transactions of $20 each, over a longer period. This is closer to the canonical whale — and what surprises studios looking at the data carefully is that the canonical whale rarely sets out to spend thousands. The amounts accumulate from many moderate-sized transactions. The typical mobile gaming whale's individual transaction is reportedly around $20.

These three player types have nothing in common except their lifetime spend total. They respond to different offers, have different retention curves, and produce different downstream value. Treating them as one segment optimizes for none of them well.

What sequence-based segmentation does

The advanced segmentation that 2026 analytics teams are building looks at the sequence of purchases, not just the cumulative total.

A player who made 5 purchases over 3 months, each around $20, with stable timing, is a systematic purchaser. They engage with the game's economy as part of their play pattern. They're likely to keep buying at the same rate unless something in the game changes their relationship with it.

A player who made 1 purchase of $100 three months ago and nothing since is a one-time buyer with whatever conversion path produced the initial purchase. They're not in the same downstream-value category as the systematic purchaser, even though both might be tagged "whale" by a lifetime-spend filter.

A player whose purchases cluster around specific in-game events — they spend during seasonal events, otherwise free-play — is an event-responsive segment. They're best monetized by event-driven offer design, not by always-on offer pressure.

A player with an accelerating purchase sequence — first month at $20, second month at $50, third month at $80 — is on a different trajectory than a player whose purchases are flat or decelerating. The accelerating player is investing more deeply; the decelerating player is showing pre-churn signals.

The sequence reveals what the cumulative total hides. Two players at $200 lifetime spend can be a clearly retaining engaged spender and a soon-to-leave former spender. The right next action for each is different. The right offer for each is different. The right value to put on retaining each is different.

Compound criteria beat single thresholds

Beyond sequence, the other shift in 2026 segmentation is toward compound criteria.

A typical single-threshold rule might be "players who spent more than $500 in their lifetime." This catches a population but mixes spenders from different periods (some recent, some lapsed) and different engagement levels (some still active, some inactive).

A compound rule does better: "players who spent more than $500 in the last 30 days AND made at least 5 separate purchases in that window AND have session activity in the last 7 days." This catches a much smaller but much more cohesive segment — players who are currently active, spending recently, and spending in a pattern that suggests an ongoing relationship with the game.

The smaller segment has higher downstream value per player and responds more predictably to interventions. The bigger lifetime-only segment is a mix that produces average outcomes from averaging contradictory behaviors.

The compound approach extends naturally. A "high-value at-risk" segment combines high spend recency with declining session frequency — these are the players a studio most wants to identify because losing them is expensive. A "developing whale" segment combines moderate spend with accelerating session frequency — these are the players to invest in retaining before they become high-value. Each compound segment has a clear intervention; each requires looking at more than one signal.

What this changes about offers

The practical implication is that offer design needs to vary by segment more finely than it usually does.

The systematic $20-transaction whale doesn't need to be shown $99.99 bundles. The conversion rate on those bundles for this segment is structurally low — they buy in $20 increments by preference, not by accident. Showing them larger bundles produces irrelevance and offer fatigue without producing transactions.

The one-time impulse buyer benefits from periodic high-perceived-value offers timed to engagement spikes. They don't have a systematic purchase pattern; the studio has to catch them in moments of higher intent and present something worth converting.

The event-responsive segment benefits from event-driven escalation. They convert during seasonal periods at much higher rates than during baseline periods. The studio's offer mix during events should be calibrated to this segment's preferences; off-event offers can be lighter.

The accelerating purchaser benefits from offers that match their growth — slightly higher tiers as their engagement deepens, premium content access, recognition of their elevated status. Showing them the same starter-tier offers they responded to two months ago wastes the relationship.

A single global offer mix optimized for the average "whale" produces decent results for none of these specifically. A segment-aware mix produces better results for all of them.

What "moving toward this" actually looks like

For studios still segmenting on lifetime payer/non-payer, the practical path forward involves a few discrete steps.

Start tracking the sequence of purchases, not just the cumulative total. Most analytics setups have this data; few teams query it routinely. The query that returns "for each player, the timestamps and amounts of all purchases in the last 90 days" is the foundation.

Define your top-priority compound segments explicitly. Five well-defined segments with clear criteria beat fifty fuzzy ones. The segments that move revenue most are usually: active high-spenders, developing high-spenders, at-risk recent spenders, event-responsive, and one-time large purchasers.

Audit which offers are being shown to which segments. Studios often discover that their offer mix is being shown roughly uniformly across segments because the segmentation that drives offer selection is coarser than the segmentation the analytics team can produce. Closing this gap — making the offer engine consume the better segmentation the analytics produces — is usually where the biggest revenue lift hides.

Run holdout tests on the segment-aware offer treatments before declaring victory. A segment that converts better in your model might be doing so because it was going to convert anyway; the right test is whether the segment-aware treatment outperforms the segment-blind treatment on the same population.

See how we approach per-segment offer treatment →

The 5%/65% concentration isn't going away — if anything it's becoming more pronounced as the broader player base spends less per capita and the most-engaged segment becomes a larger share of total revenue. The studios who understand the shape of their top 5% in more detail than "they spent a lot" are positioned to extract more from a population that already dominates their P&L. The studios who keep segmenting on the payer/non-payer line are leaving most of the available lift on the table.

Talk to us about segment-aware monetisation →