Articles

    How to Compare Affiliate and Partner Program Software for iGaming

    By Florian WrenholtAugust 6, 2026Updated October 26, 202611 min read

    In the example month below, with invented figures, €10,000 of GGR becomes €6,000 of NGR and €1,800 of commission for one partner. Then a €300 chargeback posts on the second day of the next month. Whether that partner keeps trusting the programme depends on whether the software can show where the €300 went. Most vendor demos never get that far. They show clicks, sign-ups and balances, and skip the expensive question: can the platform produce a statement that matches finance's ledger and still gives the affiliate enough detail to believe it?

    Write the commercial model down before the first demo

    A tool built for a single-brand casino on one rev-share (revenue share) plan can fall apart once a multi-brand group mixes CPA (a fixed fee per qualifying player), hybrids (CPA plus rev share), sub-affiliates and local taxes. So the buyer's first document is a map of legal entities, brands, products, currencies, player events and payment schedules. Draw it before anyone books a demo. Then the team will not buy a portal whose commission engine cannot follow deals already signed.

    Plan for the cases that break simple assumptions. Brand- and product-specific deals, cohort terms with start dates, CPA validation periods, sub-affiliate trees, payment in other currencies and players shared across brands should all be configurable without custom code for each deal. A group with 4 brands, 3 deal types and 2 payout currencies already has 24 combinations a hard-coded engine must get right.

    The test is cheap and decisive. Load three real agreements into a script: a CPA, a rev share with deductions, and a hybrid across two brands. Require the engine to calculate all three from sample events with no spreadsheet repair afterwards. A vendor that asks for two weeks of professional services to configure the hybrid has answered the question already.

    Attribution rules decide whether partners trust the numbers

    Attribution is the contract layer between marketing and payment. With vague rules, paid media, influencers, affiliates and CRM all claim the same player. Compare the whole chain: click IDs, sub-IDs, landing page parameters, referral codes and server-to-server events. The platform should keep the original source, the current source and the rule that decides who gets paid. This matters when a player signs up on mobile, deposits directly three weeks later, or moves from a content site into an app.

    Software worth comparing lets you test each attribution behaviour against those cases instead of choosing one for you:

    • last-click, first-click, fixed-source, and assisted models;
    • cross-device identity after consented login, app installs, and deferred deep links;
    • promotion codes used without an affiliate click, and organic brand search after a referral;
    • excluded operator email, paid search, internal testing, and migration across regulated entities.

    Last-click for everything is convenient and often wrong. A review affiliate can create the value long before a late coupon click takes the credit. Say a comparison site sends a player who reads two reviews, leaves, and returns 12 days later through a cashback coupon. Last-click pays the coupon site 100% and the reviewer nothing. Within two quarters the reviewer stops writing about the brand. The software should apply whatever policy you choose, consistently, and never force its own default.

    Partners see every one of these rules through the portal, and they judge the whole programme by it. A portal that shows only totals turns every attribution decision into an email thread. One that shows the event trail lets partners settle most questions themselves. Give managers a controlled dispute view per player, with the non-sensitive click, KYC, deposit and reversal events. An explained rejected commission does far less damage than an unexplained zero. Never show player identity in a dispute, and confirm that data was collected lawfully and that withdrawn consent is respected.

    In a demo, apply a restricted GEO and a custom deal, create a deep link and a sub-ID, upload a creative and request a wallet withdrawal. Count every manual handoff. With 300 active partners, one handoff that causes one ticket per partner a month means 300 tickets, about one full-time support job.

    Regulated markets add a creative layer. The library must show which assets, claims and landing pages are approved for each GEO, and stay up to date when rules change. When Great Britain's new promotion rules banned mixed-product offers in 2026, a banner saying "bet £10, get a free bet and 50 spins" became non-compliant overnight wherever it still ran. A library with no approval status and no expiry date cannot tell you which partners were still using it.

    Revenue definitions must survive a finance review

    A portal figure is only as good as the revenue definition behind it. Clicks and FTDs (first-time depositors) are simple. Rev share is not: it depends on NGR definitions that change by product, licence, market and agreement. Check how the platform treats GGR, bonus cost, taxes, payment and game-provider fees, chargebacks and jackpot contributions. Check that it applies the signed deal exactly and shows a calculation a person can read.

    Timing is where definitions break. A casino player might generate slots GGR, take free spins, trigger a chargeback weeks later and leave an unsettled withdrawal at month end. A sportsbook player places bets funded by free bets that settle next period. If the platform cannot say when revenue counts for commission or can be reversed, finance rebuilds statements outside it, and the ledger and portal drift apart. Define carryover, voided bets, duplicate accounts and manual adjustments clearly. Give every adjustment a reason, an approver, a timestamp and a history that cannot be changed.

    Here is how that reconciliation reads on one partner's month (one brand, one 30% rev-share deal with deductions):

    NGR = GGR - (bonus cost + gaming tax + payment & provider fees + chargebacks)

    Line Amount
    GGR (stakes minus wins) €10,000
    Less bonus cost -€1,500
    Less gaming tax -€1,200
    Less payment & game-provider fees -€800
    Less chargebacks settled this month -€500
    NGR €6,000
    Commission = NGR × 30% €1,800

    Now test the timing. Under this agreement, the €300 chargeback posted after month end goes on the next statement. Under a negative-carryover deal, it reduces next month's NGR. Say next month the partner's players produce €2,000 of GGR and €2,500 of deductions. NGR is minus €500, the €300 chargeback takes it to minus €800, that amount carries forward, and commission is zero. The month after, €4,000 of NGR first covers the €800, leaving €3,200 and a commission of €960. Without carryover, the same partner would have been paid €1,200 that month. That €240 gap is exactly what a partner disputes when the portal cannot show it.

    Before signing, reconcile sample monthly statements with the ledger. Include late settlements, refunds, bonus costs, fraud exclusions and a mid-term deal change. Match both the totals and the player-level logic. If the platform cannot show these lines with the same figures as the ledger, finance will rebuild statements by hand. Exposing that gap is the whole point of this comparison.

    Commission flexibility needs hard guardrails

    A flexible commission engine lets the affiliate team be creative with deals. Without checks, it creates liabilities just as easily. Compare CPA, rev share, hybrids, fixed placements, FTD bounties, sub-affiliate rev share, tiers and short-term incentives, but judge each by what it rewards.

    Goodhart's law, named after the economist Charles Goodhart in 1975, applies here: when a measure becomes a target, it stops measuring well. Pay on FTD count, and partners will deliver FTDs, including ones who never deposit again. For CPA, define a qualifying FTD as completed KYC, an approved payment method, a minimum of real-money play and no fraud flags within a stated period. Then check the payback. Say you pay a €150 CPA on players whose average NGR is €25 a month at a 40% contribution margin. It takes 15 months to earn back, before churn. If only 35% of players from that source are still active on day 90, it never does.

    The same deal design includes the safer-gambling gate. A qualifying event should exclude self-excluded customers, accounts under a safer-gambling intervention and players blocked by deposit limits. These exclusions should reverse an already accrued CPA automatically, without waiting for a manual clawback. Operators licensed in Great Britain already have to switch marketing off for a customer showing strong signs of harm, and a commission that keeps paying on that customer's play sends partners the opposite message.

    For rev share, look closely at negative carryover, how long the deal lasts, who owns the player and what is deducted. Carryover smooths out swings. But if it never ends, it drives away small publishers, so make it a deliberate commercial choice. For hybrids, be clear whether an upfront CPA is an advance, a separate reward or something you can claw back. Test generous offers before you recruit with them. They attract promo-only traffic and fake accounts, and growth turns into lost margin.

    Payments and fraud decide the programme's reputation

    Partners judge a commission plan by the date the money lands. Accurate statements earn no trust if payments arrive late or without explanation. Walk the workflow from approved balance through tax validation, payment-file release, remittance, failure handling and reconciliation, with thresholds, beneficiary validation, invoices, withholding and risk holds supported.

    Build in separation of duties. An affiliate manager approves an adjustment, finance releases the payment, and a balance frozen by risk is released only by named approvers. Surface the statuses in the portal (pending, approved, paid, failed, on hold) with non-sensitive reasons attached, because silence is what turns a five-day delay into a relationship problem.

    Fraud control is the other half of that reputation. One weak source can create hundreds of low-quality sign-ups. So link partner performance to duplicate accounts, reused payment methods, shared devices, chargebacks and bonus-only activity. Borrow a habit from experimentation teams: a number that looks surprisingly good is more often a tracking or data error than a breakthrough. A new partner converting 60% of sign-ups to FTD when the programme average is 25% needs an audit before a bonus.

    Treat fraud flags as review inputs, since a shared device can be completely innocent. Risk teams must record reasons, attach evidence, hold balances and document outcomes. Monitor by affiliate, campaign and GEO, and compare registration-to-FTD, KYC pass rate, early chargebacks and day-30 NGR. Cheap CPA gets expensive fast once failures, bonuses and churn arrive.

    Data and integration edges decide credibility after an outage

    Numbers like day-30 NGR by affiliate or KYC pass rate by campaign have to be calculated somewhere, and rarely inside the vendor's own screens. Product, finance, BI and CRM need controlled data they can reconcile and keep outside the interface. Check the APIs, export limits, webhooks, delays, retention and documentation. Confirm that the API gives you the events and fields you need without a vendor ticket.

    Require stable IDs for partner, campaign, click, registration, player reference, brand, GEO, deal version and currency, with pseudonymous player IDs and documented time zones. A time zone mismatch alone moves revenue between statements. A platform on UTC and a ledger on CET disagree about every bet placed between 23:00 and midnight UTC on the last day of the month. Daily exports may be fine for monthly payments but can miss a wave of invalid sign-ups or a broken postback for 24 hours. Match the refresh speed to the decision it supports.

    Integration work hides at the edges of the player lifecycle: KYC that changes after a deposit, chargebacks that post after approval, country changes, open bets at month end, and self-excluded users who must drop out of marketing. For each event, name the source, the owner, the expected delay, the retry rule and how it is delivered. Ask directly about duplicate, out-of-order and unmatched events. Safe retries and monitoring decide whether the numbers are trusted after an outage. For multi-brand groups, a player acquired for Brand A who then registers with Brand B needs a configurable, audited rule.

    Score vendors on a 30-day proof, then govern on retained NGR

    Turn the scripted demos into a weighted scorecard, because impressions in the room favour the best presenter. A new operator may care most about launch speed and ready-made templates. A mature group cares most about reconciliation, fraud, data access and control across legal entities. Give each area an owner from growth, finance, risk, engineering or compliance. Treat roadmap features as missing until the contract sets terms, timing and remedies. Compare setup fees, minimums, per-conversion and payout fees, export charges, custom work, contract length and exit rights. A low headline fee often hides paid reporting or API add-ons that appear once you are locked in.

    Then run a 30-day trial on controlled sample data, never live balances. Start with links and exclusions. Then registration, KYC, deposit, bonus, wager, settlement and reversal events. Then three real deals, all the way to a reconciled statement. Finally onboarding, creative approval, payment holds and escalation. Set pass rules in advance: no unexplained ledger difference, a documented fix for every failed event and an owner for every support path in production.

    All this rigour costs money, and some operators can skip it. A single-brand operator in one GEO with 20 partners on a flat CPA can launch on a simple tracker in a week and spend the saved months on acquisition. That shortcut works while the deals stay simple. It breaks at the first rev-share deal with deductions or the second brand. The cost of fixing statements by hand grows with every partner, and moving historical attribution later is much harder than choosing well now.

    Once live, the platform keeps getting brands, deals and GEOs the trial never saw, so manage it on those terms. Hold a monthly control review with affiliate operations, finance, risk, compliance and BI. Judge quality by retained NGR, not FTD count. Check promotional materials by GEO and approval status. Look at held commissions, chargebacks, duplicate patterns and complaints. Exclude restricted, self-excluded, underage and other ineligible users from marketing and commission wherever the law requires it. Never reward messages that push deposits, claim gambling is a source of income or target vulnerable people. Each jurisdiction, licence, product and channel changes the legal picture, so local lawyers should check terms, tax, marketing and payout duties before launch.

    The decision rule is short. If the trial month ends with every ledger difference explained to the euro and every held commission linked to a reason, sign. Then settle in the contract who owns data, support, configuration and change control. If either test fails, the platform has just shown you its first year of disputes.