Showing free-spin winnings as withdrawable cash rather than as bonus credit cut the share of players who re-spent those winnings on further spins by 18% at 9pm, according to operator-side data from a six-week test across three UK-licensed brands. The change is purely presentational — same winnings, same wagering, same withdrawal rules — but the 9pm reading is the one that matters, because that is when the behavioural gap between "cash" and "bonus" labels is widest. Below is what the test actually measured, where the effect holds, and where the numbers get thin.

What changed, and what didn't

The intervention was a label swap. Free-spin winnings that had already cleared their wagering requirement were displayed in the cash balance with a pound sign and a "withdrawable" tag, instead of sitting in a separate bonus meter that read "£4.20 bonus" until the player manually converted it. No terms moved. The 35x wagering on the spin winnings stayed at 35x. The 24-hour expiry on unconverted balances stayed at 24 hours. Maximum conversion per player per day stayed at £50.

That matters for how you read the result. This is not a bonus-terms story, and it is not an argument that looser wagering drives retention — the opposite, if anything. It is a story about salience. A balance labelled "bonus" reads to most players as house money with a catch attached, and house money gets staked again. A balance labelled "cash" reads as theirs, and theirs gets withdrawn or sat on.

The test ran from 14 April to 26 May 2025, split roughly 50/50 within each brand, with a control group seeing the old bonus-meter display. Combined sample was 41,600 players who triggered at least one free-spin award during the window. The headline 18% figure is the relative reduction in the proportion of converted winnings re-staked on slots within 24 hours of conversion, measured at the 9pm hour specifically.

Why 9pm is the number to watch

Aggregate figures across the whole day showed a 9% reduction — real, but modest, and the kind of gap that could plausibly be sampling noise on a smaller sample. The effect was not flat across the day. It clustered hard in the evening.

Hour band	Relative reduction in re-staking
06:00–12:00	3% (not significant)
12:00–17:00	7%
17:00–21:00	12%
21:00–23:00	18%
23:00–02:00	14%

The 9pm peak lines up with when free-spin awards are most commonly triggered — operators in the test pushed most awards between 19:00 and 22:00 — and with when players are most likely to be playing on a phone, in a low-attention context, with a balance small enough that the label is the main thing distinguishing it from spendable cash. At 10am, a player with £6.40 in converted winnings and a coffee in front of them behaves differently from the same player at 9pm with the TV on.

The mechanism: labelling, not liquidity

It would be easy to read this as a liquidity effect — cash is easier to withdraw, so of course less of it gets re-staked. But the test held liquidity constant. In both arms, the player could withdraw the converted amount at any point; the control group simply had to tap "convert to cash" first. Same money, same access, one extra tap and a different word on the meter.

The gap persisted after controlling for that tap. Players in the treatment arm who did convert and then re-staked did so at similar stake sizes to the control group — the difference was in whether they re-staked at all, not how much. That points at the label doing the work, not friction.

There is a familiar parallel here. Research on casino loyalty schemes has long shown that players treat "bonus" and "cash" as psychologically distinct even when the expected value is identical, and the UK's 2020 ban on credit-card gambling deposits produced a similar pattern: removing a funding route changed behaviour far more than the equivalent restriction on stake size would have predicted. Labels and rails shape play more than the arithmetic underneath them.

What it means for the re-spin economy

Free spins are, functionally, a re-staking machine. The award exists to put a player in front of a slot with a balance they did not deposit, on the reasonable bet that some proportion will keep going once the spins are done. Operators have spent years tuning the mechanics — spin value, expiry, max conversion — to push that proportion up.

This test suggests a chunk of that tuning is fighting the label. If 18% of the 9pm re-stake behaviour disappears the moment the winnings say "cash", then the bonus meter was doing measurable work as a retention device, and removing it is a genuine cost. The operators in the test will have modelled that. The question they are sitting with is whether the 18% is worth the compliance and trust upside of showing players their money as money.

The compliance angle nobody put in the press release

The timing is not accidental. The Gambling Commission's ongoing focus on bonus transparency — specifically the Consumer Protection and ADR requirements that came into force in 2022, and the subsequent consultations on "financially vulnerable" customers — has pushed operators toward clearer balance presentation. A meter that says "bonus" when the money is, in every legal sense, the player's after wagering is cleared sits awkwardly against that direction.

The 9pm figure is also the one that will interest the Commission most. Late-evening, mobile, low-attention play is exactly the context its guidance on identifying harm flags. An 18% reduction in re-staking at that hour is a small but clean data point that presentational clarity moves behaviour in the direction regulators want. That is not the same as saying the test was designed to please anyone — it wasn't, by the operators' own account — but it does mean the result will travel further than a typical A/B test.

There is a counter-reading worth taking seriously. If showing winnings as cash reduces re-staking, it also reduces the entertainment value of the free spin for players who wanted to keep playing. A player who converts and withdraws £4.20 has had less play per award. Whether that is a harm reduction or a product degradation depends on which side of the counter you stand.

The open question: does the effect decay?

The obvious unknown is durability. Six weeks is long enough to rule out novelty as the whole story, but not long enough to know whether players habituate. If the treatment arm's players learn over three months that their free-spin winnings always show as cash, do they drift back toward control-group behaviour? The test didn't run long enough to say, and none of the three brands has published a follow-up.

The second unknown is whether the effect survives contact with the broader bonus ecosystem. This test isolated free-spin conversion. Most of these players also hold deposit-match bonuses, reload offers and loyalty credit, all of which still display as bonus. If the "cash" label only applies to one small balance among several, players may simply recalibrate which meter they ignore. The 18% could be an artefact of contrast — cash stands out precisely because everything around it says bonus.

The third is the one that should worry anyone extrapolating too hard: the effect was measured on re-staking within 24 hours. It says nothing about total session length, deposit behaviour, or lifetime value. An operator could see an 18% drop in re-spins at 9pm and a completely flat revenue line if those players simply came back the next evening instead. The test's authors know this. The number is real; what it's worth is still an open bet.