Affiliate Program KPIs: The 12 Metrics to Track and What Each One Tells You
Most affiliate dashboards answer one question: how much revenue did partners bring in this month? It is a fair question, but it is the last number in a long chain. By the time revenue drops, the cause happened weeks earlier: partners stopped activating, a landing page stopped converting, a tracking change started dropping sales, or your best partner quietly moved budget to a competitor.
The affiliate program KPIs worth tracking are the ones that show you where in that chain the problem sits. This post covers twelve of them, grouped by stage, with the formula for each, what a bad reading usually means, and a short monthly scorecard you can build from data your platform already has. It goes deeper on the measurement side of our affiliate program audit checklist.
Why top-line affiliate revenue is a lagging metric
Partner revenue is the product of several things multiplied together: how many partners you have, how many of them are active, how much traffic each active partner sends, how well that traffic converts, and how much of the resulting revenue survives refunds and reversals.
Tracking each factor separately is what turns "the program is down" into "activation fell after we changed the onboarding email" or "one partner type converts at half the rate of the others".
The 12 affiliate program metrics, by stage
Any figures below are illustrations to show the arithmetic, not benchmarks. What counts as healthy depends on your price point, category and partner mix, so the most useful comparison is your own program over time.
Stage 1: Recruitment
1. Qualified applications per month
Formula: applications that match your target partner profile, divided by month.
What it tells you: whether your recruitment is reaching the partners you actually want. A large number of applications with a low share of qualified ones usually means the program page attracts everyone and convinces no one in particular.
2. Approval-to-first-click time
Formula: median days between a partner being approved and their first tracked click.
What it tells you: how much friction sits between "yes" and "live". Partners who wait days for approval, then cannot find a working link or a usable asset, often never come back. If this number is long, the fix is usually onboarding, not recruitment.
Stage 2: Activation
3. Activation rate
Formula: partners who drove at least one sale within a set window after approval (for example 60 or 90 days), divided by partners approved in the same cohort.
What it tells you: whether the program gives new partners enough momentum to start. This is the metric most programs skip, and often the one that explains the most. A program that recruits well and activates badly has a leak, not a traffic problem. We covered the causes in most affiliate programs do not have a traffic problem.
4. Active partner ratio
Formula: partners with at least one sale in the last 30 or 90 days, divided by all approved partners.
What it tells you: the real size of your program. Watch whether the active core is growing, holding or shrinking month by month.
Stage 3: Partner economics
5. Earnings per click (EPC)
Formula: approved commission paid to partners, divided by clicks they sent, in the same period.
What it tells you: what promoting you is worth to a partner compared with the other programs they could send the same traffic to. Track it per partner type, because a single program-wide EPC hides the partner groups that are earning very little. The structure behind it is covered in affiliate commission structure: models, examples and how to choose.
6. Click-to-sale conversion rate
Formula: attributed sales, divided by partner clicks.
What it tells you: whether the traffic partners send and the page it lands on fit each other. Compare affiliate conversion with your own organic traffic to the same page. A large gap points to intent mismatch, weak partner landing pages, or tracking that drops sales between click and checkout.
7. Effective commission rate
Formula: total commission paid, divided by partner-sourced revenue net of refunds.
What it tells you: what the program really costs, after tiers, bonuses, coupons and reversals. It is often different from the rate on your program page, and it is the number to hold against your margins.
Stage 4: Revenue and customer quality
8. Net partner-sourced revenue
Formula: revenue from partner-attributed customers, minus refunds, chargebacks and cancelled orders.
What it tells you: the revenue that actually stayed. Gross partner revenue can grow while net revenue falls, which is a sign that some partners are sending customers who do not stick.
9. Reversal rate
Formula: commissions reversed, divided by commissions created, per partner and per partner type.
What it tells you: customer quality and, sometimes, abuse. A partner whose reversal rate sits well above the rest of the program deserves a closer look at the traffic they send and how they describe your offer.
10. Partner-referred customer retention
Formula: share of partner-referred customers still paying (or still buying) after a set period, such as 3, 6 or 12 months, compared with customers from your other channels.
What it tells you: whether you are paying for customers worth having. For subscription businesses this matters more than the first sale. If partner customers churn faster than your average, the right response is usually to change who you recruit or what event you pay on, not to cut rates across the board.
Stage 5: Program health
11. Revenue concentration
Formula: share of partner-sourced revenue that comes from your top five or top ten partners.
What it tells you: how fragile the program is. Heavy concentration is common and not a problem by itself, but it means one partner leaving can move your quarter. It is a reason to manage those relationships deliberately and invest in the next tier down.
12. Partner churn
Formula: partners who were active last period and drove no sales this period, divided by partners active last period.
What it tells you: how many productive partners you are losing, which top-line revenue hides until it is too late. A partner who goes from steady sales to none rarely tells you why. The common reasons are in why affiliates stop promoting your program.
A worked example: reading the numbers together
This is an illustration with made-up round numbers, to show how the metrics combine. It is not a real program.
A SaaS program approves 120 new partners in a quarter. Its revenue from partners is flat compared with last quarter, and the team assumes it needs more partners.
- Activation rate: 12 of the 120 drove a sale within 90 days, so 10%. Nine in ten new partners never start.
- Approval-to-first-click time: a median of 9 days, because approvals are reviewed manually once a week and the welcome email has no tracking link in it.
- EPC by partner type: content partners earn noticeably more per click than the coupon partners, because coupon traffic converts well but mostly on existing customers who were already buying, and those orders are later excluded.
- Revenue concentration: the top five partners bring in about 70% of partner revenue.
Read together, the story is not "we need more partners". It is: the program loses most new partners in onboarding, its revenue depends on a handful of relationships, and part of its partner mix produces reversals. Recruiting another 120 partners into the same onboarding would mostly produce another 108 inactive accounts. The higher-value fixes are a same-day approval and a welcome sequence with a working link, a closer relationship with the top five, and clearer rules for coupon partners.
If you want a rough sense of what gaps like these might be costing, the free revenue leak calculator gives a monthly estimate from inputs you already have.
A one-page monthly scorecard
You do not need a data team to track this. Most affiliate platforms export partners, clicks, sales and commissions, and your billing system holds refunds and retention. A spreadsheet with one row per month and these columns is enough to start:
- Qualified applications
- Partners approved, and median days to first click
- Activation rate for the cohort approved 90 days ago
- Active partners and active partner ratio
- Clicks, sales and click-to-sale conversion
- EPC, overall and for your two or three main partner types
- Net partner-sourced revenue and reversal rate
- Effective commission rate
- Top-five revenue concentration
- Partners who went quiet this month, by name
Read the scorecard as a set of trends, not single readings. Add customer retention for partner-referred cohorts once a quarter, because it needs time to mean anything.
The last column matters more than it looks. A named list of partners who went quiet is the only metric on the sheet that tells you who to call.
Common measurement mistakes
- Counting total partners as a success metric. A growing list of inactive accounts is not growth. Report active partners instead.
- Using pending commissions in EPC. Pending numbers include orders that will be refunded or reversed, which makes the program look better to you than it is to partners.
- Averaging across partner types. Content, paid media, coupon and influencer partners behave differently. A single average hides which group is working.
- Ignoring tracking gaps. If affiliate-attributed sales in your platform do not roughly match what your own billing system shows for the same customers, every other metric is built on a shaky base. Reconcile the two before trusting trends.
Where this fits in a full program review
Metrics are how we find the leak, but not where we stop. In a program review we rate six elements: recruitment, activation, payout structure, retention, tracking and creative. The KPIs above map onto those six, which is why a weak reading points to a specific part of the program to fix.
Frequently asked questions
What are the most important KPIs for an affiliate program?
If you track only a few, track the activation rate (the share of approved partners who drive a first sale), earnings per click for your partners, partner-sourced revenue net of reversals, and revenue concentration in your top partners. Together they tell you whether partners start, whether promoting you pays, whether the revenue is real, and how fragile the program is.
What is a good conversion rate for affiliate traffic?
It depends on your price point, your offer and the partner type, so a number from another program tells you little. The useful comparison is internal: affiliate traffic against your other channels to the same page, and each partner type against the others. If affiliate visitors convert far below your own organic visitors on the same page, look at intent mismatch and landing pages before blaming partners.
How do you calculate earnings per click (EPC) for an affiliate program?
Divide the total commission you paid or approved to partners in a period by the number of clicks they sent in the same period. Calculate it per partner and per partner type as well as for the whole program, and use approved commissions rather than pending ones so reversals do not inflate it.
How often should affiliate program metrics be reviewed?
Review a short scorecard monthly and a deeper set quarterly. Monthly is often enough to catch a partner going quiet or a tracking break before it costs a full quarter. Customer quality metrics such as retention and refund rates need longer windows, so read them quarterly against cohorts of customers acquired in the same period.
Should affiliate KPIs include customer lifetime value?
Yes, if you can tie customers back to the partner who referred them. Without it you optimise for first sales and can end up paying the most to partners who send customers who refund or churn. Compare the retention and value of partner-referred customers with customers from your other channels, by partner type where the numbers allow.
Not sure which of your numbers is the real problem? The Affiliate Program Audit pulls your recruitment, activation, partner economics, tracking and customer quality data into one view, finds where the program leaks, and hands back the fixes ranked by revenue impact. Express Audit $297, Full Audit $1,950.
Request Audit