The Affiliate Dashboard Audit: Which Numbers Actually Predict Profit and Which Just Feel Good

Open your affiliate dashboard and you will probably see clicks, conversions, total commissions, and a chart that goes up and down. Every one of those numbers is accurate, and most of them will not tell you what to do next. A page can get plenty of clicks and earn nothing, and a month of high commissions can shrink after refunds are processed.

This audit ranks the numbers by one test: does the number change a decision? If a figure would not make you rewrite a page, move a link, drop a program, or leave something alone, it can be checked monthly or ignored. The affiliate marketing metrics that matter are the few that pass this test, and all of them can be calculated from the dashboard you already have.

This article assumes your tracking is already in place. If it is not, start with how to track affiliate clicks and conversions, then come back here to read what the tracking produces.

Why a full dashboard can still leave you guessing

Dashboards are built to report activity, not to support decisions. They show totals because totals are easy to produce, and they show them across the whole account because that is the simplest grouping. A total across ten pages and three programs cannot tell you which page to fix.

Reading the dashboard well means turning totals into ratios (earnings per click, conversion rate), splitting them by page and source, and comparing them over time. A single total tells you what happened. A ratio compared across pages tells you where to act.

There is a second problem: the dashboard reports what was tracked, not what was kept. Commissions can sit pending, get reversed, or be adjusted after a return. Reading the number as income before those things resolve overstates your results.

Vanity numbers to stop checking daily (and the one job each still does)

Some numbers are fine to glance at but poor guides to action. Checking them daily mostly produces reactions to normal fluctuation. Each still has one legitimate use.

• Total clicks. Clicks measure attention on your links, not money. Their job is to be the denominator in earnings per click and conversion rate. A rising click count with flat commissions means the ratio is falling, which is worth knowing, but the count alone is not.
• Pageviews and sessions. Traffic is the raw material, and it also says nothing about whether visitors want what you recommend. Use it to calculate how often visitors click an affiliate link: outbound clicks divided by sessions on that page.
• Total commissions. This is the number that feels best and informs least. It is useful for cash flow and tax records. It mixes pending, approved, and reversed money, so it does not describe performance.
• Impressions and rankings. A ranking tells you whether people can find a page. Use it to diagnose a traffic drop, and do not treat it as a measure of profit.
• Number of links or programs. More links is not more income. Count them only when checking that each page has a link in the right place.

If you currently look at the dashboard every morning, switch to one scheduled review. The audit below takes an hour and replaces dozens of small checks.

Earnings per click: the single number that compares every page fairly

Earnings per click (EPC) is commissions earned divided by clicks sent. It puts a page with 5,000 visits and a page with 300 visits on the same scale, because the traffic volume cancels out.

Here is an illustration with invented numbers. Page A gets 1,200 sessions and 180 outbound affiliate clicks, and earns $90 in the period. Page B gets 300 sessions and 60 clicks, and earns $75.

• Page A: $90 ÷ 180 clicks = $0.50 per click, or $0.075 per session.
• Page B: $75 ÷ 60 clicks = $1.25 per click, or $0.25 per session.

Page A has four times the traffic and earns less per visitor. That points to a decision: Page B deserves more traffic, internal links, and a closer look at why it works, while Page A needs diagnosis.

Some details affect whether the number is trustworthy:

• Check the definition. Many networks report EPC per 100 clicks, and some count only unique clicks. Calculate your own figure from raw clicks and commissions so every page is measured the same way.
• Use the same period for both numbers. Commissions are sometimes dated by the click and sometimes by the sale. For programs with long cookie windows, a sale can land weeks after the click, so a short period understates EPC. Use a longer window, such as 60 or 90 days, for stable figures.
• Treat small samples as undecided. A page with a few dozen clicks and one or two sales will produce a wild EPC. Rank pages by EPC only when each has enough clicks to make a single sale a minor part of the result.

Track both EPC and earnings per session. EPC says how good the offer and the click are. Earnings per session also reflects whether the page gets people to click at all.

Conversion rate by page and by traffic source, not as a site-wide average

A site-wide conversion rate is an average of very different situations. It will move when your traffic mix changes even if no individual page changed. Calculate it per page and per source instead.

Conversion rate here is conversions divided by clicks. Splitting a page’s performance into two stages makes the diagnosis clearer:

1. Click rate: outbound clicks ÷ sessions. This measures whether the page persuades visitors to move toward the merchant.
2. Conversion rate: sales ÷ clicks. This measures whether the visitors who clicked then bought.

The two stages point to different problems.

• Low click rate, normal conversion rate. The page is not getting enough people to the link. Look at link placement, the clarity of the call to action, and whether the search intent behind the page matches what you recommend. A page that ranks for a research question may not belong to a buying audience.
• Healthy click rate, low conversion rate. Visitors are interested but something breaks after the click. Check that the link still goes to the right product page, that the price or offer on the merchant’s page matches what you described, and that the audience you attract can afford or needs the product. The page may promise more than the merchant’s page delivers.
• Both low. Review whether the page should recommend this product at all.

Then split by source. Organic search, email, and social traffic often convert differently on the same page, because the visitors arrive with different intent. If your dashboard or analytics tool supports sub-IDs or campaign tags, use them. If a source converts far below the others, compare it to the rest before changing the page, because the fix may be to stop sending that source to that page.

Reversals, refunds and pending commissions: the gap between reported and real income

Reported income and real income are different figures, and the gap between them is information about your content.

Pull three numbers for the same period:

• Pending: commissions recorded but not yet approved. Programs hold these for different lengths of time, so read each program’s terms for its approval window.
• Approved or paid: money you can count.
• Reversed or refunded: commissions removed after a return, cancellation, chargeback, or fraud check.

The reversal rate is reversed commissions divided by total commissions earned for the sales in that period. Calculate it per program and, if the dashboard allows, per page. Calculate it by month of sale, not by the month the reversal was processed, so the rate lines up with the sales that produced it.

What a high reversal rate usually signals depends on where it sits:

• High on one page, normal elsewhere. The page probably sets expectations the product does not meet. Reread what you claim, and compare it to what the merchant delivers.
• High across one program. The product itself may have a refund problem, or the program’s attribution may be unreliable. Weigh that before sending more traffic.
• High for one traffic source. That source may bring buyers who are less sure of their decision, such as impulse clicks from social posts.

Pending balances need patience, not action. Judge a month’s performance only after most of its commissions have been approved or reversed. Until then, treat the figure as an estimate, and avoid making decisions based on the most recent weeks.

Revenue concentration: how much of your income rides on one page or one program

Concentration is the share of income that comes from your biggest sources. Calculate two figures: top page revenue ÷ total revenue, and top program revenue ÷ total revenue. Add the top three of each if your income is spread unevenly.

This number does not say anything is wrong. It measures exposure. If one page produces most of your commissions, a ranking drop, a deleted product page, or a program rule change affects most of your income at once. If one program produces most of it, a change in its commission rate or cookie window does the same.

There is no standard threshold, so set your own. A reasonable approach is to decide in advance what share from one page or program would make you uncomfortable, and write it down. When the audit shows you are past that line, the response is to build a second source, not to cut the strong one. That could mean a new page targeting a related query, a second program for the same audience, or a different income type. Programs that pay monthly can reduce reliance on one-time sales, and how to build recurring affiliate income covers that route.

Also watch the opposite pattern. If your income is spread evenly across dozens of pages that each earn very little, the issue is not concentration but weak earnings per click. Ranking pages by EPC will show it.

A one-hour audit routine: what to pull, what to compare, what to write down

Run this monthly, or quarterly if your volume is low. Set a timer, and keep notes in a single spreadsheet so each audit can be compared with the last.

Minutes 0 to 10: pull the data. Export, for the last 90 days: sessions per page from your analytics tool, outbound affiliate clicks per page, and commissions per page and per program, split into pending, approved, and reversed. Also export the same figures for the 90 days before that.

Minutes 10 to 25: build the page table. One row per page, with these columns:

Page Sessions Outbound clicks Click rate Commissions EPC Sales Conversion rate Change vs prior 90 days

Sort by EPC. Mark pages with too few clicks to judge as “undecided” instead of ranking them.

Minutes 25 to 35: split by source. For your top five pages, compare conversion rate by traffic source where the data exists.

Minutes 35 to 45: check reversals. Calculate reversal rate per program and flag any page or program that stands out from the rest. Note the pending total separately, and exclude it from your performance figures.

Minutes 45 to 50: check concentration. Write down the top page share and top program share, and compare them with the limit you set.

Minutes 50 to 55: check the trend. Compare each figure against the previous 90 days. Look at direction on EPC, click rate, and reversal rate, not only on total commissions. If you have a year of data, compare the same period a year earlier, because seasonal buying patterns can look like a decline or a gain when they are neither.

Minutes 55 to 60: write down findings. Record three things: what changed, what you think caused it, and what you will do about it. Without this step, next month’s audit has nothing to compare against.

Turning findings into a fix list, and when to leave a page alone

Rank the findings by how much income is at stake. A short list you finish beats a long one you abandon.

A workable order:

1. High traffic, low EPC. These pages have the most room for improvement, since fixing them multiplies across existing visitors. Diagnose with the click rate and conversion rate split above.
2. High reversal rate. Correct claims that overstate a product, add a clear description of who the product is not for, and consider dropping a program whose refund rate stays high.
3. High EPC, low traffic. These pages convert well but are hard to find. Add internal links from related pages, update the content, and check whether the target query has more search demand than you are capturing.
4. High concentration. Plan one new page or program aimed at reducing dependence on your top source.

Some pages should be left alone:

• Pages with too few clicks to judge. Changing a page based on three sales is guessing. Wait for more data.
• Pages changed recently. If you edited a page in the last few weeks, let the new version collect a full period of data before judging it.
• Pages with a seasonal dip. If the same drop appeared at the same time last year, the cause is probably the calendar.
• Stable earners. A page with steady EPC and a low reversal rate does not need a rewrite because other pages do.

Change one thing per page at a time and record the date. If you rewrite the introduction, move the link, and swap the product in one edit, you will not know which change moved the numbers.

Your first audit sets the baseline, and the second shows whether your fixes worked. Pick your three weakest pages from the audit and work through the fix process in the case study on fixing affiliate content that stopped converting.