Affiliate Campaign Metrics: Connecting Clicks to Approved Revenue

A campaign can attract more clicks while producing less approved revenue. That is why affiliate reporting needs to follow the full journey from impression to payout instead of celebrating one improving percentage. The original Stars Partners article highlighted CTR, CPC, EPC, landing conversion, FTD, retention and LTV. This guide turns that list into a working measurement system with explicit denominators, comparable cohorts and clear limits on benchmarks. The goal is a report that explains the next decision, not the biggest dashboard you can build.
Define every metric before comparing campaigns
CTR is clicks divided by impressions, expressed as a percentage. Google notes that a good CTR depends on what you advertise and on which network. It is an attention signal, but on its own it does not show that the user understood the offer or became a paying customer.
CPC is the relevant media spend divided by the corresponding clicks. EPC is the selected earnings divided by the same clicks. State whether those earnings are provisional, approved or paid, because the same campaign can show a different EPC once conversions are reviewed.
Landing conversion needs a named event and a denominator. A visit-to-registration rate and a click-to-registration rate differ when some clicks never load the page. If the difference helps diagnosis, label both instead of calling every percentage a conversion rate.
FTD means first-time deposit, and an FTD rate still needs its denominator. FTD divided by registrations measures a different stage from FTD divided by clicks. Specify whether the deposit is recorded, qualified or approved by the advertiser, and whether each person is counted only once.
Keep dates and currencies aligned. A click-date cohort, a conversion-date total and a payment-date total answer different questions. Otherwise conversion delay and currency conversion can make a comparison misleading even when every formula is correct on its own.
Read the CPC and EPC spread as a partial margin signal
When approved EPC exceeds CPC on the same clicks, earnings cover the media cost. That is a good sign, but it is not yet profit. Creative production, infrastructure, fees, staff and later adjustments sit outside both figures.
Take an illustrative cohort: 1,000 clicks, $200 media spend and $300 approved affiliate earnings. CPC is $0.20 and EPC is $0.30, leaving $100 for other costs. These numbers are invented for teaching, not a Stars Partners result or a market benchmark.
If another $120 of costs belongs to that cohort, the same positive spread per click comes with a $20 loss. Show the cost breakdown next to the result. Media margin and full campaign profit should be separate lines, not one unlabeled headline.
When the spread narrows, do not jump to blaming the bid. The cause may be a higher CPC, lower conversion, a different offer mix, delayed approval or a larger adjustment. Change the part the evidence points to instead of editing targeting, creative and landing all at once.
Connect landing behaviour with qualified first deposits
A landing page should deliver the promise of the ad and make the next step clear. Check loading errors, language, navigation and whether users actually reach registration. Better copy cannot fix a broken link, and a faster page cannot turn the wrong audience into the right one.
The Milliseconds Make Millions study, presented by Google, linked a 0.1-second improvement in mobile site speed to an 8.4% conversion increase in retail and 10.1% in travel. These are findings from other sectors, not an iGaming forecast. Measure speed in your own funnel instead of copying those percentages into a revenue plan.
The step from registration to approved FTD shows where money is lost. If registrations rise while approved FTDs fall, check the audience promise, product access, qualification criteria, event collection and approval timing. The pattern tells you where to look, but it does not name the cause.
Compare one controlled segment first. Where possible, keep the same source, GEO, offer, device type and cohort age. A combined total can hide a change in traffic mix that looks like a landing improvement but actually comes from different visitors.
Compare retention and LTV through mature cohorts
Retention needs a defined return event and a cohort start. Reopening the app, making another deposit and staying commercially active are different actions. Write down the rule before Day 1 or Day 30 appears on a chart.
The original article cited AppsFlyer retention for casino and betting apps: on Android 28.15% on Day 1 and 12.47% on Day 30, on iOS 25.05% and 7.56%. The English version also cited Adjust averages across all categories: 26% and 7%. We could not find a matching dated dataset with those filters for this update, so treat these figures as historical source claims, not current targets.
For LTV, explain whose value you measure. Player deposits, operator net revenue, affiliate commission and contribution after costs are different quantities. State the period and separate measured value from a forecast.
Compare cohorts of equal age. A cohort acquired yesterday has no observed Day 30 yet, and a blank should never be quietly counted as zero. Mark incomplete periods and show how many users stand behind each rate, because a small cohort makes percentages jumpy.
Build reporting around a question and an owner
The original Stars Partners article presented full-cycle support, analytics, funnel refinement and direct-access offers as its services. That is the partner's own statement, and we have not audited it. The framework below is editorial guidance for any team evaluating performance.
One report can connect spend, clicks, registrations, qualified FTDs, approved earnings and cohort value. Keep source, creative, offer, device, GEO and the relevant dates available as dimensions. Name the person responsible for each question so that a chart leads to a next action.
| Pattern | Useful next check |
|---|---|
| CTR rises, approved EPC falls | Audience promise and downstream quality |
| Clicks stable, loaded visits fall | Routing and page availability |
| Registrations rise, qualified FTD falls | Qualification, product fit and reporting delay |
| Provisional EPC rises, approved EPC falls | Approval and adjustment reasons |
| New cohorts look weaker | Equal cohort age and traffic mix |
Before explaining a sudden movement, reconcile your systems:
- time zones;
- event definitions;
- duplicate rules;
- missing postbacks;
- recent configuration changes.
Otherwise a tracking incident can turn into an unnecessary creative overhaul.
Daily monitoring catches incidents, while cohort decisions often need a longer observation period. Agree which movements need investigation and what reporting delay is normal. Base the rules on your own data and business risk: there is no universal threshold for stopping or scaling.
Questions about the affiliate dashboard
Does a high CTR mean high-quality traffic?
No, it only shows clicks relative to impressions. Check downstream behaviour and the approved outcome before drawing conclusions about quality.
Is a positive EPC minus CPC the campaign's profit?
Only if all relevant costs are included, and these two metrics usually leave many out. Track media margin and total contribution separately.
Can published retention figures become our KPI?
Not without the original period, filters, definitions and a comparable audience. Your own cohort data with agreed rules is a stronger basis.
Which number should be checked first after a drop?
First confirm that data collection and reporting still work, then find the earliest confirmed break in the journey. Find the cause before changing the whole campaign.
Make the chain explain the next decision
A useful dashboard joins attention, conversion, approval and player value through consistent definitions. Stars Partners' original selection of metrics is a good starting point, while precise benchmarks need more context than the source gave. Keep the denominator visible, compare cohorts of equal age and separate provisional earnings from final ones. A metric earns its place when it helps the team find a real problem and choose an action whose effect can be checked later.







