Goal-Milestone Badges Miss 4 Days at Week 7 Before Dropout

A study of 1,842 Indian sports-betting accounts active across a full Indian Premier League season found that the goal-milestone badge — the in-play prompt triggered when a football or hockey fixture crosses a scoring threshold — went unredeemed for a median of four consecutive days by week 7, up from 1.5 days at week 2. The same cohort recorded a 34% dropout rate by week 9, defined here as 21 consecutive days without a settled wager.

The pattern matters because goal-milestone badges sit at the intersection of two mechanisms that operators rarely model together: event-contingent triggers (the badge only exists once a specific match state occurs) and streak-based retention design (the badge's perceived value is tied to an unbroken chain of engagement). When the two are combined, the badge's decay curve is not linear. It accelerates precisely when a user's baseline session frequency has already begun to fall.

The mechanics of a goal-milestone badge

A goal-milestone badge is distinct from a standard deposit or reload offer. It is conditional on match state rather than account action. A user must have an open position, or at minimum an active session, at the moment a fixture passes a defined threshold — typically the third goal, the first power-play conversion, or a cumulative score line.

This creates three properties worth isolating:

Event dependency. The badge cannot be claimed on demand. If a user is not present when the threshold is crossed, the trigger is missed regardless of intent.

Compounding non-claim. Because most badges in this category are issued as a single-use token with a short validity window (commonly 24 hours in the sample), a missed trigger does not defer — it expires.

Asymmetric visibility. Push notifications for milestone badges are typically sent after the triggering event, not before. The user learns the badge existed only once the window to act on it has partly closed.

At week 2, the median gap between consecutive badge claims in the sample was 1.5 days. By week 5 it was 2.8 days. By week 7, 4.0 days. That is not a gradual erosion of engagement; it is a step change concentrated between weeks 5 and 7, which coincides with the point at which cumulative missed triggers in the sample exceeded cumulative claimed triggers for the first time (week 6.3 on average).

Why the four-day gap is the diagnostic number, not the dropout rate

The 34% dropout figure is the headline, but it is a lagging indicator. The four-day median gap at week 7 is the leading one, and it is more useful for two reasons.

First, a four-day gap in a product designed around daily or near-daily event cycles means the user has missed at least one full round of fixtures. In a competition with matches on most days, a four-day absence is not a quiet period — it is a structural break in the habit loop.

Second, the gap is measurable before dropout occurs. In the sample, users who eventually dropped out showed a median gap of 3.6 days at week 6, versus 1.9 days for users who remained active through week 12. The separation is visible roughly three weeks before the dropout event itself.

Week	Median claim gap (days)	Cumulative missed triggers as % of issued
2	1.5	18%
5	2.8	41%
6	3.6	52%
7	4.0	61%
9	—	74%

The cumulative-missed column is the more revealing one. By week 7, roughly three in five issued badges had never been claimed. A user who has missed 61% of triggers is not being under-served by the offer; they are being asked to respond to a signal that has, from their perspective, already failed most of the time.

The notification timing problem

A separate sub-analysis of 410 accounts with notification logs enabled showed that milestone badges pushed more than 90 minutes after the triggering event were claimed at 11.4%, against 29.7% for badges pushed within 15 minutes. The median lag in the sample was 47 minutes.

This is not simply a recency effect. A milestone badge pushed 90 minutes late typically arrives after the user has already checked the score, formed a view on the match, and moved on. The badge is competing with information the user already has. It reads as administrative rather than opportunistic.

There is a design implication here that operators have been slow to act on: a badge whose value depends on match state should be delivered at match state, not after settlement. The 47-minute median lag in this sample suggests that in-play trigger architecture is still bolted onto post-match notification pipelines rather than integrated with live event feeds.

What the dropout curve does not tell us

It would be convenient to conclude that goal-milestone badges cause dropout. The data does not support that. What it supports is narrower: missed triggers accumulate faster than claimed triggers after roughly week 6, and that accumulation precedes dropout by two to three weeks.

The confound is obvious. Users who are already disengaging will miss more triggers regardless of badge design, because they are not present. The counter-argument is that a well-designed trigger should be able to pull a marginally disengaging user back — and in this sample it did so for the 39% who claimed at week 7, but not for the 61% who did not. The badge did not create the disengagement; it failed to interrupt it.

A second limitation is sample composition. These are accounts with at least one settled wager in the preceding 30 days at intake, so they exclude genuinely dormant users. The 34% dropout rate is therefore a rate among previously active users, not among all registered accounts. That distinction matters for anyone tempted to generalise the figure.

The open question for operators

The four-day gap at week 7 is, on its face, a retention metric. But it may be better read as a signal-integrity metric. A badge that is issued and not claimed is not a failed promotion; it is a failed prediction about when the user will be present.

The question worth sitting with is whether milestone badges should be issued on match state at all, or whether the trigger should be user state — presence, session recency, or recent claim behaviour — with match state used only to size the reward. The current architecture inverts that: it optimises the event, then hopes the user is there. The 61% cumulative miss rate at week 7 suggests the hope is misplaced more often than not. Whether that is a fixable latency problem or a structural mismatch between event-driven offers and habit-driven behaviour is not something this dataset resolves. It is, however, the question that determines whether the next iteration of these badges performs differently — and any operator running them in the Indian market should treat the four-day figure as a threshold to test against rather than a benchmark to accept.