Why SaaS affiliate programs fail, and how to find your cause

SaaS affiliate programs fail for six reasons, all inside the program: affiliates who are approved and then ignored, a commission losing the comparison with a competitor's, recruiting by volume, tracking dropping credit, terms inviting refunds and abuse, and no check against the billing ledger.

No study I could find measures a failure rate, so the figures below come from two datasets. The first is Reditus's analysis of 10,596 affiliate-months.

The second is the TrackRev affiliate benchmarks, which compile public reports from Impact, PartnerStack, Awin, Rakuten Advertising and CJ Affiliate. According to Reditus, over 90% of affiliate-months produce no paid referral.

In the TrackRev benchmarks for Q2 2026, only 18 of every 100 affiliates make a sale in the first 30 days.

The sections below give the first check for each of the 6 causes, from the 18% median activation rate to the 120-day dispute window.

Sources: getreditus.com/blog

What does the data say about programs that stall?

A stalled program is a program where a few partners carry the revenue and the rest of the roster stays quiet. The table lists 5 measures from the two datasets, 3 from Reditus and 2 from the TrackRev benchmarks.

The SaaS affiliate program benchmarks for 2026 cover the wider set.

By these numbers a normal program is a few productive partners and a long quiet roster.

The top 10% of affiliates bring 71% of revenue in the median program, Reditus found the top 20% of active affiliates produced 80% to 87% of monthly revenue, and only 18 of every 100 affiliates make a sale in the first 30 days.

Past referrals keep paying after a partner stops sending clicks, so the month's click count undercounts a program.

A program fails when the partners who would carry the program never show up or never get started, or when the offer or the tracking drives partners away.

Why do SaaS affiliate programs fail? Six causes

Six causes explain a stalled program. Each cause has a first check in the symptom table later in the post.

1. Why do approved affiliates never get activated?

Activation is the share of new affiliates who make a first sale within 30 days. According to the TrackRev affiliate benchmarks, top-quartile programs reach 31% and typically run a welcome sequence, ready-made creative and a first-week check-in.

The median program sends an approval email and a dashboard link and reaches 18%, and the bottom quartile approves and never follows up, reaching 6%.

First, send the first message the day someone is approved, with the tracked link, the one page the link should point to, the terms in one line, 100 words of ready copy and a name that replies within a day.

The guide to creating an affiliate program has the week-one sequence. For example, a 31% rate turns 100 approved affiliates into 31 sellers, against 18 at the median and 6 in the bottom quartile.

Second, list the affiliates approved in the last 60 days and count how many have a single click.

2. Why does the commission lose the comparison?

An affiliate choosing what to feature compares what a referred customer pays. On a $99 monthly plan, four common structures pay very differently, as the table shows.

The one-time payout is an eighth of the 20% recurring one ($29.70 against $237.60).

The benchmarks put the median B2B SaaS program at 20% for 12 months and B2C SaaS at 15% for 6 months, and 39% of programs pay fixed-term recurring commission, 33% pay once at the sale and 18% pay for life.

At an 80% gross margin, 20% for 12 months takes 25% of a customer's year-one gross profit ($237.60 of $950.40) and 30% takes 37.5%.

Lifetime commission is the other trap: Reditus calls lifetime commission the hardest term to walk back once affiliates have signed up on it.

I would start with a fixed term, extend the term for partners who earn it, and write down what your two closest competitors advertise before settling on a rate.

3. Why does recruiting by volume fail?

Recruiting by volume is the habit of approving every applicant. First, count the roster: the bottom half of affiliates brings 2% of revenue, so open approval mostly grows the part of the roster that does not matter.

Reditus ties its best case, $30K MRR in 3 months, to aggressive recruiting plus letting affiliates bid on the brand name in paid search.

Second, a hand-built list of 20 names beats an open sign-up page, and the list has 3 sources:

• customers who already post or write about your product
• authors of comparison and "best tools" pages that rank for your category
• owners of newsletters and channels whose readers match your buyer, found through Product Hunt launches or Hacker News threads about your category

Decide early about coupon and deal sites. A publicly posted code reaches every shopper about to check out, so the site collects commission on customers already at your checkout.

Keep coupon sites out for the first 6 months or issue each site a code owned by one partner alone, as the coupon code tracking guide explains.

4. Why does tracking drop the credit?

Three things drop credit, and one program can have all three: the window, the browser and renewals.

How does a short attribution window lose sales?

An attribution window is the number of days a click can still earn credit for a sale.

In the benchmark data 28% of programs use 30 days or less, 41% use 60, 24% use 90 and 7% use 180 or more, and Reditus calls 90 days a reasonable default for considered purchases.

TrackRev's workspace data shows why: the 75th percentile of days from first click to first paid charge is 21, the 90th is 48.2 and the 95th is 73.5.

A 30-day window therefore cuts off roughly 10% to 25% of buyers, and the affiliate who sent the buyers sees a click and no sale.

For example, a $99 plan with a 20% commission pays $237.60 over 12 months, and a missed first sale forfeits the whole $237.60.

How does Safari delete affiliate cookies?

Safari Intelligent Tracking Prevention is the WebKit policy that shortens the life of script-written cookies.

WebKit's Intelligent Tracking Prevention policy says Safari deletes cookies created in JavaScript after 7 days without the visitor interacting with your site, and caps the cookies at 24 hours when Safari detects link decoration, meaning click IDs added to a URL and read by script on the landing page.

Cookies set in an HTTP response get the 7-day cap when Safari detects CNAME or IP cloaking, and the policy names no general cap otherwise.

So a 60-day window stored by a script can be 7 days long for a Safari visitor who does not return within a week, about 12% of the intended window.

For example, a Safari visitor who clicks on day 1 and buys on day 9 earns the affiliate no credit, so the $237.60 commission on a $99 plan never starts. First, set the cookie from your server.

Second, test in Safari with a purchase on day 8. Finally, compare the result with Chrome in 2026.

Why do renewals lose affiliate credit?

A renewal is a paid invoice with no browser involved. A thank-you-page pixel fires once, at signup, while a subscription renews as an invoice announced by Stripe Billing with an invoice.paid event.

A $99 plan produces 11 renewal events after signup, which hold $1,089 of the $1,188 year-one revenue, or 92%.

The client_reference_id on a Stripe Checkout Session is one place to carry the referral ID, and how affiliate tracking works shows where each link in the chain breaks. First, carry the referral ID into checkout.

Second, credit every invoice.paid event to the original click. To size your own loss, compare 90 days of paid invoices for affiliate-referred customers with the sales the affiliate dashboard shows.

5. Why do terms and payouts invite disputes?

Terms invite disputes when commission is paid before the refund window closes. The result is chasing money from someone who has spent the money: $19.80 per refunded sale at 20% on a $99 plan.

A commission hold is a delay between a sale and the payout, and a hold that matches the refund window removes most of the loss.

Disputes run longer: Stripe says card networks typically allow 120 days from the original payment, and an opened dispute debits the amount plus a fee from your balance.

Commission has to reverse on its own when charge.refunded or charge.dispute.created arrives, and clawback handling covers partners already paid.

Self-referral, an affiliate buying through the affiliate's own link, earns commission on a sale that involved no marketing.

The Federal Trade Commission (FTC) says affiliates must disclose the relationship near the recommendation and that advertisers need a reasonable program to train and monitor affiliates, so put both rules in the first message (see affiliate program compliance for the FTC and GDPR), along with a request for an IRS Form W-9 or Form W-8BEN before the first payout.

6. Why does nobody check the program against the ledger?

The ledger is the billing record of what customers actually paid, and signups and clicks say little beside the ledger.

Track affiliates with a click in the last 30 days, revenue per active affiliate, affiliate share of new MRR, 12-month retention of referred customers against everyone else, and commission paid as a share of affiliate revenue.

TrackRev's workspace benchmarks show why to measure partners beside other channels. At $99 a month, 1 retained customer is worth $1,188 a year.

Per Q2 2026 data, the affiliate and partner channel has a median click-to-paid rate of 3.9% and a 12-month LTV multiplier of 1.4x, against 2.4% and 0.9x for paid search, and affiliate customers wait a median 9.6 months before first churn against 7.1.

The figures are medians across workspaces, so yours may sit well above or below the medians.

Which cause should you check first?

The symptom table is a lookup from what the dashboard shows to the likely cause and the first check. Start with the row matching today's dashboard.

For example, almost no clicks after week 2 points to cause 1, and an affiliate dashboard with fewer sales than Stripe Dashboard shows for the same customers points to cause 4.

Several first checks need only 2 exports: 90 days of paid invoices and the affiliate roster.

A commission hold shorter than the refund window points to cause 5, where each refunded sale costs $19.80 at 20% on a $99 plan in 2026.

Run 1 check, fix the cause the check finds, and wait a full billing cycle before moving to the next row.

When is an affiliate program the wrong channel?

An affiliate program is the wrong channel when the break-even count cannot be reached. Before fixing anything, check that the economics can work.

Take a $99 plan at an 80% gross margin, a 20% commission, software at $29 a month, and 20 hours a month of your time valued at $50 an hour.

Eighteen counts customers paying at the same moment.

Five partners who each send 1 customer a month that sticks reach 20 active customers in month 4, ignoring churn, and if you add 30 customers a month overall, 5 from partners is one in 6.

Three situations make the channel a poor fit:

• checkout is not self-serve, so no payment event exists to credit
• the gross margin cannot absorb 15% to 30% for 6 to 12 months
• you cannot name 20 people who write or talk about your category

What does TrackRev cover and what does it leave to you?

TrackRev is SaaS affiliate software for indie hackers and small SaaS teams, with link tracking and revenue attribution built in. We built the Affiliate Program in TrackRev for the mechanical causes and left the human ones to you.

First, TrackRev handles the mechanical part of all 7 rows in the table. Second, you keep the decisions: the welcome message, the rate and term, the recruiting list, the pixel install, the hold and the monthly review.

TrackRev syncs revenue from 6 billing processors in 2026, and the 2 paid plans, Indie at $29 a month and Growth at $299 a year, cover 1 to 10 workspaces.

Every paid plan includes all of the features above, from $29 a month for Indie or $299 a year for Growth, with no limit on tracked revenue and 0% fees.

The free plan covers link tracking only (50 links, 1,000 tracked events a month), so a program needs a paid plan. For a marketplace of affiliates to recruit from, see Reditus vs TrackRev.