When an affiliate site’s revenue softens, the instinct is to publish more. Write more reviews, chase more keywords, add more pages. But most sites we might look at are not short on traffic or content. They are quietly losing a slice of the revenue they already earn, at a dozen small points between the click and the payout. Those leaks rarely show up on the affiliate dashboard, because the dashboard only reports the money that made it through. The money that didn’t is invisible by definition. Finding it is less about growth and more about plumbing.
The click-to-commission path has more joints than you think
Every affiliate sale travels a longer path than it appears. A reader clicks a link, a redirect fires, a cookie gets set, the merchant records the session, the reader buys within the attribution window, the network reconciles the sale, and eventually a commission posts. Each joint in that pipe can leak. A link cloaker that adds latency loses impatient clicks. A merchant that shortened its cookie window from 30 days to 24 hours quietly strips out every delayed purchase. A network that “de-dupes” against the merchant’s own email traffic hands your sale to someone else.
The trouble is that these failures look identical to “nobody bought.” Your analytics show the outbound click; the commission simply never arrives, and you have no natural moment where the absence announces itself. So you have to go looking on purpose.
Start with the mechanical failures because they are the cheapest to fix and the most common. Broken and redirected merchant links are the single most underrated leak in this business. Merchants restructure their sites, retire product SKUs, and change domains constantly, and every one of those changes can turn a monetized link into a dead end or an unattributed redirect. On a site with a few thousand outbound links, it is entirely normal to find that some meaningful share now point somewhere that no longer pays you.
Rule of thumb: if you have never run a full link audit, assume 5 to 15 percent of your outbound affiliate links are broken, redirected, or pointing at a program you are no longer credited for. That is revenue you have already earned the traffic for.
Attribution leaks are worse than they look
Mechanical leaks cost you clicks. Attribution leaks cost you sales you actually generated. These are harder to see because the reader buys, the merchant ships, everyone is happy, and yet the commission lands somewhere other than your account.
The usual culprits are predictable once you know to look for them. Coupon and cashback sites that sit in the last-click position steal credit at the checkout page, after you did the work of convincing the reader. A merchant’s own retargeting emails can reset the attribution to “email” if the network honors that channel over affiliate. Readers who click on mobile and buy on desktop fall out of any cookie-based system entirely. And a surprising amount of loss is just self-inflicted: internal links, staff clicks, and your own newsletter competing with your own articles for the same cookie.
You will not close all of these. Some are structural to how a given network attributes. But you can quantify them, and quantifying changes the decision. If you can show that a particular merchant’s attribution rules cost you a third of your credited sales, that is no longer a nuisance. That is a reason to negotiate, to move volume to a competitor, or to route around the merchant’s own checkout coupons.
Stale content leaks slowly and constantly
The third leak is the least dramatic and often the largest in total. Content ages, and aging content converts worse every month it sits untouched. A “best of” roundup written two years ago may still rank and still draw traffic, but if half its recommended products are discontinued, out of stock, or now priced above a competitor, the clicks it sends convert at a fraction of their original rate. The traffic looks fine. The revenue per visit has quietly collapsed.
This is a leak precisely because the page still works well enough not to demand attention. Nobody flags a page that earns something. You have to find it by comparing earnings per session across pages of similar age and intent, and looking for the ones that have drifted down.
A simple order of operations
You do not need a data team to find these. You need a repeatable pass, run in order of payoff.
| Symptom | Likely leak | First test |
|---|---|---|
| Clicks up, commissions flat | Broken or redirected links | Full outbound link audit against live merchant URLs |
| Sales you expected never post | Attribution loss at checkout | Compare your click data to merchant-side sale timestamps |
| Traffic steady, revenue per visit falling | Stale content and dead products | Rank pages by earnings per session; refresh the bottom of the list |
| One merchant underperforms peers | Cookie window or de-duplication rules | Read the actual program terms, then compare to a competing program |
The sequence matters. Links first, because they are cheap to fix and the loss is total on every affected click. Attribution second, because it changes negotiating posture even when you can’t fully solve it. Stale content third, because it is ongoing work rather than a one-time fix, and it pays best once the faster leaks are sealed.
The larger point is that the affiliate dashboard is a revenue report, not a diagnostic tool. It tells you what you kept, never what you lost on the way. A publisher who audits links quarterly, watches earnings per session by page, and reads program terms before assuming a merchant is honest will keep a meaningfully larger share of the revenue their traffic already produces, without publishing a single new word. Before you spend the next quarter making more, spend a week finding out how much of what you already make is quietly draining away.