Free Spins Economics: What Bonus Rounds Really Cost and Return

Retention · 2026-07-11 · 8 min read · By CROCO Games

Free spins are the industry's default promo currency, and most operators can't price them. Cost math, breakage, segment response, abuse patterns, and how to measure real uplift.

Free spins are the default promotional currency of online casino — welcome packages, reload offers, churn-back campaigns, tournament prizes. And yet ask an operator what 50 free spins on a given title actually cost, and what they return against a holdout group, and the answer is usually a shrug plus a CRM dashboard showing participation, which is not the same thing. This article does the arithmetic, summarises what behavioural research says about why free spins work (the answer is not "because they're free"), and lays out a measurement approach that separates real uplift from self-congratulation.

First, separate the two things called "free spins"

The term covers two unrelated objects, and conflating them muddles every conversation:

  1. In-game free spins — the bonus round inside a slot, triggered by scatters, priced into the game's RTP by the studio. The player "pays" for these through the base game; they cost the operator nothing beyond the game's normal math. Their cadence and volatility are design decisions on the PAR sheet.
  2. Promotional free spins — spins granted by the operator as a marketing cost: a defined stake, on a defined game, with defined wagering terms. These are the ones with an invoice attached, and the ones this article prices.

The cost math, honestly

The expected raw cost of a promotional free-spins grant is:

cost = number of spins × spin value × game RTP

Fifty spins at €0.10 on a 96% title have an expected winnings value of 50 × 0.10 × 0.96 = €4.80. That is what leaves your pocket in expectation if winnings were withdrawable cash. They almost never are, which is where the two adjustments come in:

So the honest unit economics of a "50 free spins" offer is usually €1–3 of true expected cost — which is why free spins beat cash bonuses on cost-efficiency, and why their perceived value (the player mentally prices them near face stake value) is the real product. You are buying perception at a discount. That asymmetry is the entire business case.

Why they work — the research is not what you think

The intuitive story is "players love free things." Taylor's 2017 series of three experiments tested this directly and found something more useful: free-spins features increased players' preference for a machine, but the driver was the audiovisual celebration and elevated win frequency during the feature — not the "free" label itself. A paid-for feature with the same sensory and win-frequency package produced the same preference. In lab conditions, players like what free spins feel like, not what they cost.

Two adjacent findings complete the picture. Chapman's 2019 lab work showed that displayed balance behaves like "house money": players with elevated balances bet larger and prefer the machine that got them there — a promotional spins win temporarily creates exactly that state. And Murch's 2024 study on real slot machines found that wins, bonus outcomes and even losses-disguised-as-wins all lengthen the pause after the outcome — the moments players savour. Promotional spins that land on a game with a strong celebration layer buy more of those savoured moments per euro.

The practical translation for operators: the game you attach spins to matters as much as the count. Fifty spins on a flat, celebration-poor title deliver less perceived value than twenty-five on a title with a rich feature presentation, at half the cost. Pick promo titles for their sensory density and win frequency, not their headline count arithmetic. (These are lab findings on preference and arousal, not guarantees of deposit behaviour — treat them as design direction, not revenue math.)

Segment response: where spins actually move numbers

Blanket free-spin campaigns are how promo budgets die. The segment logic, from field experience:

Segment Response to FS offers Right play
New depositors High — spins reduce first-session risk and teach the game loop Spins on a high-hit-frequency, low-volatility title; the goal is a long, comfortable first session (Day-2 retention is the KPI)
Healthy actives Low incremental — they were coming anyway Minimal spend; save budget. This is where holdout tests are most humbling
Churn-risk (lapsing) Moderate — a reason to return, if the game is familiar Spins on their historical favourite title, not your promo-of-the-month
Bonus hunters High participation, negative value Exclude ruthlessly; see abuse below

The common thread: spins work when they remove a reason not to play (risk for novices, inertia for lapsers) and waste money when they subsidise play that would have happened anyway.

Abuse patterns to design against

Free-spins offers attract systematic exploitation, and the patterns are stable across markets: multi-accounting to farm welcome spins; low-stake wagering grinders who complete play-through on minimal-volatility games; arbitrage between spin-grant value and low-margin wagering routes; and bonus-to-withdrawal pipelines timed across operators. Defences are equally standard — device and payment fingerprinting, wagering contribution weightings, game exclusions on the wagering path, caps on convertible winnings — but the design principle matters more than the toolset: cap the downside per account and make the grind not worth the hourly rate. A €50 max-conversion cap turns a farming operation into a hobby.

Measuring uplift without lying to yourself

Participation is not uplift. Redemption rate is not uplift. The only honest measure is incremental behaviour against a randomized holdout: take the eligible segment, withhold the offer from a random 10–20%, and compare deposits, active days and net revenue over the following 30+ days. Run it per segment, because the blended number hides the fact that healthy actives typically show near-zero incrementality while new-depositor and lapsed segments carry the campaign.

Three discipline rules: pre-register the metric (net revenue after bonus cost, not turnover); run the holdout long enough to capture wagering completion; and charge breakage savings honestly — a campaign that looks cheap because nobody redeemed it also moved nobody. If your CRM tool cannot randomize holdouts, that is a tooling gap worth fixing before the next campaign, not after. Our operator playbook covers how CROCO supports promo tooling — free-rounds APIs, per-title configuration and the math documentation to price offers correctly — and the team can walk through campaign design via contact.

Frequently asked questions

How much do 50 free spins actually cost an operator? Expected winnings are spins × stake × RTP — about €4.80 at €0.10 stakes on a 96% game. After wagering requirements and breakage, the true expected cost is typically €1–3. The perceived value to the player is far higher, which is the point.

Do free spins work because players love free things? Lab research says no — the preference driver is the audiovisual celebration and elevated win frequency of the feature, not the "free" label. A paid feature with the same package produced the same preference. Choose promo titles for sensory density, not just spin count.

Which players should get free-spin offers? New depositors (to de-risk the first sessions) and lapsing players (attached to a familiar title) show the clearest incremental response. Healthy actives mostly bank the value without changing behaviour, and bonus hunters are net-negative — exclude them.

How do I know if a free-spins campaign actually worked? Only with a randomized holdout: withhold the offer from 10–20% of the eligible segment and compare net revenue and active days over 30+ days. Participation and redemption metrics measure activity, not incrementality.

Key takeaways