LTV (Customer Lifetime Value) in iGaming: Definition, Formula and How Operators Read It
Customer Lifetime Value (LTV) is the total revenue, typically NGR, an iGaming operator can expect a player to generate over the entire lifecycle. It is the metric that turns every short-term decision (CPA, bonus spend, channel allocation, segment treatment) into a long-term economic…
TL;DR
Customer Lifetime Value (LTV) is the total revenue, typically NGR, an iGaming operator can expect a player to generate over the entire lifecycle. It is the metric that turns every short-term decision (CPA, bonus spend, channel allocation, segment treatment) into a long-term economic question. LTV is also one of the most-modelled and most-misused metrics in the industry, because it depends on assumptions about retention and monetisation that very few operators measure cleanly.
Formula
In its simplest form:
Two common variants exist. Historical LTV looks backward: how much NGR has this player generated to date. Projected LTV looks forward: how much NGR will this player generate over the remaining lifecycle, based on observed behaviour and modelled retention. Both are valid for different decisions.
A typical historical LTV aggregation looks like this:
Warehouse-level aggregationSELECT
player_id,
MIN(bet_date) AS first_bet_date,
MAX(bet_date) AS last_bet_date,
SUM(stakes - winnings
- bonus_cost
- payment_fees
- gaming_tax) AS lifetime_ngr
FROM bets
WHERE bet_status = 'settled'
GROUP BY player_id;Projected LTV is more interesting and harder. Most operators use a model that combines observed early behaviour (deposits, bets, sessions in the first 7-30 days) with statistical or ML-based retention curves to estimate forward NGR over a defined horizon (commonly 12 or 24 months). The quality of the model determines the quality of every CPA payback decision the operator makes.
Why it matters in iGaming
iGaming acquisition is expensive. CPA in mature regulated markets often runs in the high double digits to low triple digits in euros. The only way to know whether that spend is profitable is to compare it against the NGR each acquired player will generate over time. LTV is the answer to that question, which is why every commercial decision in iGaming acquisition implicitly references it, whether the team uses the term or not.
Different teams read LTV differently:
- Marketing uses LTV-to-CPA ratio per channel as the core signal of channel quality.
- CRM uses LTV per cohort to size lifecycle programmes and prioritise reactivation spend.
- Finance uses cohort LTV as a forecasting input for revenue planning.
- Executives watch LTV trends as a top-level signal of acquisition quality and platform health.
LTV is also one of the few metrics where modelling sophistication directly affects business outcomes. Operators with naive LTV models (e.g. average historical NGR multiplied by an arbitrary lifetime) systematically misjudge channel quality. Operators with proper cohort-based projected LTV models reach correct conclusions earlier and adjust acquisition spend faster.
Common mistakes and how teams get LTV wrong
Confusing historical LTV with projected LTV. Historical LTV tells you what already happened. Projected LTV tells you what to expect. Most acquisition decisions need projected LTV. Reporting historical LTV as if it were forward-looking is the most common error in CPA payback analysis.
Using GGR instead of NGR. GGR-based LTV ignores bonus cost, payment fees and tax, all of which are real and material. NGR-based LTV is the honest version. Operators that use GGR-based LTV systematically overestimate channel profitability.
Aggregating across very different cohorts. A welcome-bonus cohort has very different LTV from an SEO-driven cohort. Blended LTV hides which channels are working. Cohort-level LTV by acquisition source is the only useful version.
Ignoring the time horizon. 12-month LTV, 24-month LTV and "true" lifetime LTV are different numbers. Most operators settle on 12 or 24 months as the standard projection horizon. Comparing across time horizons without alignment produces wrong conclusions.
Treating heavy-bonused early activity as durable. Early NGR from welcome-bonus play often does not project linearly to lifetime value. Naive models that extrapolate from first-week NGR overstate the LTV of bonus-driven cohorts and understate the LTV of organic ones.
Ignoring discount rate for long horizons. A euro of NGR in month 24 is worth less today than a euro of NGR in month 1. Most LTV models for acquisition decisions apply some discount, even if conservative. Treating distant NGR at face value distorts payback timing.
Healthy ranges and benchmarks
LTV varies dramatically by vertical, market, channel and player cohort. Directional patterns:
- Casino-led operators tend to deliver higher per-player LTV than sportsbook-led operators because of higher per-session monetisation and steadier engagement.
- VIP cohort LTV can be one to two orders of magnitude above casual base LTV, with very heavy concentration of total NGR in a small share of players.
- Mature regulated markets typically deliver higher LTV than emerging markets, both because of higher purchasing power and because of more durable retention patterns.
- Direct and SEO traffic tends to deliver higher LTV than aggressively bonused acquisition channels, often justifying lower acquisition volume with much better unit economics.
- Bonus-driven cohorts often show high early NGR followed by sharp drop-off, producing 12-month LTV that is much lower than first-month NGR would suggest.
These are directional patterns based on industry observation, not strict benchmarks. Cross-operator LTV comparisons are rarely useful without alignment on horizon, NGR definition and cohort segmentation.
How Gamblitude handles LTV
In Gamblitude, LTV is exposed as a governed Metric in two main forms: Historical LTV (cumulative actual NGR per player) and Projected LTV (12 or 24-month forward forecast). Both are computed at the semantic layer with explicit conventions for NGR definition, cohort attribution and horizon. Channel-Specific LTV, Cohort LTV and Segment LTV variants coexist as separate Metrics so each team can work from the right view for their decision.
Predictive LTV is computed by ML models trained on iGaming behaviour patterns, available as a per-player Attribute. CRM teams use it to build dynamic Lists of high-value early-cohort players for premium treatment, and Marketing teams compare LTV to CPA at channel level for ongoing optimisation. Insight Radar surfaces meaningful drift in cohort LTV trajectories before they affect quarterly results.
FAQ
In iGaming, the two terms are often used interchangeably. Some industries reserve LTV for the discounted, profit-margin-adjusted version of lifetime revenue. iGaming operators usually treat LTV as cumulative NGR per player over a defined horizon, with optional discount and margin adjustments depending on use case. Documenting the convention matters more than picking the canonically correct definition.
NGR. GGR-based LTV ignores bonus cost, payment fees and gaming tax, all of which are real costs of generating that revenue. NGR-based LTV is closer to actual contribution to the business. Operators that report GGR-based LTV often do so to flatter channel performance metrics; the practice is misleading.
Most operators use 12 or 24 months as the standard projection horizon. Shorter horizons (3 or 6 months) are useful for fast iteration on acquisition channels but understate true value. Longer horizons (36 months or true lifetime) produce more complete numbers but require more model confidence at the long tail. The right horizon depends on the decision being made.
Often dramatically. Direct and SEO channels typically deliver higher LTV per acquired player than aggressively bonused channels. Premium content affiliates often outperform comparison-site traffic. Cohort-level LTV by channel is one of the most useful inputs to budget allocation decisions.
Because welcome-bonus structures attract a mix of genuine new customers and bonus-arbitrage players. The latter generate strong first-week activity (driven by bonus play-through requirements) and then disappear. Naive LTV models extrapolating from early activity miss this drop-off and overstate cohort value. Cohort-aware models that observe drop-off curves directly produce more accurate projections.
Further reading
Found a mistake or want a term added to the iGaming Glossary? Let us know.
Browse the complete glossary or see how governed definitions work across dashboards, reports, alerts and AI answers.