The 9 Affiliate Program KPIs Worth Tracking (And the Vanity Ones to Skip)
A working dashboard for merchants — activation rate, effective commission cost, EPC, fraud share, concentration risk — with healthy ranges and the metrics that waste your attention.
Affiliate dashboards produce numbers effortlessly; deciding which numbers deserve a monthly look is the actual work. Here are the nine that answer real merchant questions — is this channel growing, is it profitable, is it safe — plus the vanity metrics that only feel informative.
Growth: is the channel getting bigger?
1. Referral revenue. The headline: revenue from affiliate-attributed orders per month. Track its share of total revenue too — for stores that invest in the channel, a share in the single digits climbing toward the teens over the first year is a common, healthy shape.
2. Active affiliate count. Not signups — partners with at least one referral this quarter. This is the program's true size. A hundred recruits with eight active is an eight-partner program with a mailing list.
3. Activation rate. Of newly approved affiliates, how many produce a first referral within 30 days? This measures your onboarding, not your partners: if it's low, the first promotion isn't easy enough (the activation fix). Rough banding: under a fifth activating is a problem; above a third is strong.
Economics: is it profitable?
4. Effective commission cost. Total commissions ÷ referral revenue, including the customer discounts on affiliate coupons — the stacking cost most stores forget. If your nominal rate is 10% but coupon discounts push the true cost to 19%, this metric is where you find out. Judge it against your contribution margin, not against habit.
5. Earnings per click (EPC). Commission earned per tracked click, the metric affiliates themselves use to compare programs. A healthy EPC is your best recruiting stat; a weak one usually points at the store (conversion, pricing) rather than the partners.
6. Refund/reversal share of referrals. Affiliate traffic should refund at roughly your store's normal rate. A partner whose referrals reverse at multiples of baseline is sending mismatched audiences — or worse (the fraud patterns).
Health: is it safe and durable?
7. Concentration. Revenue share of your top partner and top three. Concentration is natural — every program has a power curve — but when one partner is most of the channel, their departure is a revenue event. Past a half from one partner, recruiting insurance beats optimizing anything else.
8. Held-referral share and review latency. What fraction of referrals lands in fraud review, and how fast do you clear it? Rising held-share flags abuse or an over-tight ruleset; latency above a couple of days quietly tells partners you don't trust them. The queue exists to be emptied.
9. Payout punctuality. Binary and brutal: did every eligible partner get paid on the promised day? Twelve for twelve is a recruiting asset; eleven for twelve is a reputation problem. (The payout workflow that makes this automatic.)
The vanity list
Skip, or at least never decide based on: raw click totals (bots and accidents inflate them; EPC already contains the signal), signup count (see activation), social reach of applicants (audience ≠ influence), and cumulative all-time revenue (it only ever goes up; trends live in monthly views).
The monthly ritual
A program this size needs twenty minutes a month: scan the nine numbers, write one sentence about the biggest mover, and pick one action — a recruiting push if active count stalls, an onboarding fix if activation sags, a margin conversation if effective cost creeps. Single-action months compound; ten-initiative months evaporate.
Most of these numbers should come straight off your dashboard — referral revenue, statuses, held counts, and payout history are all first-class objects in Affiliate Factory WP, and the demo shows the reporting view with 90 days of sample data if you want to see the shape before wiring up your own.