Your network dashboard reports conversions. It cannot tell you which article produced them, which link position performed, or which topics are worth more of your time. That gap is where most affiliate sites lose their ability to improve.
The three questions worth answering
- Which pages earn? Not which get traffic — which produce revenue.
- Which links get clicked? Position, wording and format all matter, and the differences are large.
- Which topics deserve more coverage? The answer is usually not the one you'd guess.
Everything below exists to answer those three.
Sub-IDs: the most underused feature in affiliate marketing
Nearly every network supports appending a custom parameter to your links — variously called SubID, sid, u1, or similar. Whatever arrives at the network comes back in your reports.
Use it to encode the page:
https://network.example/track?id=123&subid=vpn-comparison-post
Now your commission report tells you the sale came from your VPN comparison article rather than merely "somewhere on your site". Encode position too if you want more detail — vpn-comparison-intro versus vpn-comparison-table.
This costs nothing, takes minutes, and is the difference between guessing and knowing.
Click tracking on your own side
Network reports only show clicks that arrived. To see what happens on your pages, track outbound clicks yourself.
Most analytics platforms can record outbound link clicks as events, either automatically or with a small snippet. What you want out of it:
- Click-through rate per page — what share of readers click any affiliate link
- Clicks per link position — first paragraph versus comparison table versus conclusion
- Clicks per format — inline text link versus button versus table row
The patterns are consistent enough to be worth knowing: links inside comparison tables typically outperform links in prose, and links above the fold outperform those below it — but the magnitude varies enough by audience that measuring your own is worthwhile.
Link redirects
Rather than pasting raw network URLs into content, route them through your own domain — /go/product-name redirecting to the affiliate URL.
The benefits are practical:
- Change a destination in one place when a program's URL changes, rather than editing forty articles
- Count clicks server-side, unaffected by ad blockers
- Cleaner links that don't expose tracking parameters
- Survive program changes — if you switch from one program to another, one redirect update covers every mention
Two cautions. Redirects must still carry rel="sponsored" nofollow on the originating link, and you should block /go/ in robots.txt so crawlers don't index redirect URLs. Cloaking affiliate links to disguise them from users violates most network terms — the goal is maintainability, not concealment.
Metrics that actually matter
Revenue per thousand visitors (RPM). The single most useful number. It normalises for traffic and lets you compare a small high-intent article against a large informational one directly.
Conversion rate per program. Two programs in the same category can convert very differently. If one converts at a third the rate of another, the higher headline commission may be worth less.
Click-through rate per page. Low CTR on a high-traffic page means the content isn't reaching the point of decision, or the links are badly placed.
Reversal rate. Track how many commissions get clawed back. A program with a 30% reversal rate is worth much less than its headline figures imply, and you won't notice without looking.
Time to first conversion. How long after publishing an article does it earn? This tells you how long to wait before judging new content.
A minimal setup
You don't need expensive tooling:
- Analytics installed, with outbound click events configured
- Sub-IDs on every affiliate link, encoding the source page
- Redirects through your own domain for anything you'll mention repeatedly
- A monthly spreadsheet pulling network earnings against page traffic to compute RPM
- Search Console, to see which queries bring the traffic that converts
That's a couple of hours to set up and it turns your content strategy from intuition into arithmetic.
What to do with the data
The most common finding surprises people: a small number of pages produce nearly all the revenue, and they're rarely the highest-traffic pages.
When you find them, the response is not to publish more of everything. It's to publish more of that — the same topic, adjacent comparisons, related alternatives — and to improve the pages already earning. Doubling the conversion rate on a page that already earns is usually easier than finding a new page that earns at all.