The Affiliate Report Your CFO Will Actually Read: Six Numbers That Prove the Program Pays

The affiliate manager's monthly report usually says something like: 340 clicks, 62 code uses, $8,400 in attributed revenue, 12 new partners. The CFO reads it, nods, and puts it in the folder with last month's. Nothing in it answers the only question finance has, which is whether the program makes money once you count everything it costs.

That question is answerable. It takes six numbers, most of which you can pull from the Reveshare dashboard and a Shopify export in under an hour a month. The point of this post is to name them, show where each comes from, give you a sense of what healthy looks like, and warn you about the way each one lies if you let it.

The report at the end fits on one page. That is deliberate. A one-page report gets read and a six-page one gets filed.

1. Net attributed revenue

What it is. Revenue from orders attributed to affiliates, after refunds and after the customer discount. Not gross order value, not "sales influenced", and not anything that includes an order later returned.

Where it comes from. Reveshare records commission per order as it clears and reverses it on refund, so the attributed order list already reflects reversals. Take the period's attributed orders, subtract refunded amounts, and use the net order value after the discount code, not the pre-discount subtotal. Cross-check the total against Shopify's orders filtered by the discount codes in use.

What healthy looks like. There is no universal range, because it depends entirely on program size. What matters is the trend and the share of store revenue. For a program that has been running a year, affiliate net revenue typically sits somewhere between a few percent and a fifth of store revenue. Below that, the program is a side project. Above it, concentration risk becomes a real question, which is number four.

The trap. Reporting gross. Gross attributed revenue includes the discount the customer received and the orders that came back, so it can overstate the real number by a fifth or more. If finance later reconciles it against the ledger, the gap destroys the credibility of every other number in your report.

2. Effective cost of sale

What it is. Everything the program cost, divided by net attributed revenue. Everything means commission paid or accrued, the customer discount given on attributed orders, the tool subscription, and any product seeding or bonuses. Express it as a percentage.

Where it comes from. Commission and discounts come from the same attributed order list. Tool cost is your Reveshare invoice. Seeding is whatever you shipped at cost.

Compare it with blended CAC. Take your paid channels for the same period: ad spend divided by new customers acquired through them. Then compute the affiliate equivalent: total program cost divided by new customers acquired through affiliates. Put them side by side. This single comparison is what makes the report worth reading.

Commission + discount + tools

A program paying 15% commission plus a 10% customer discount on attributed orders has a cost of sale around 25% before tool cost. Whether that is good depends on your contribution margin and what paid channels cost you for the same customer. For many Shopify brands in 2026, it compares well with paid social, but the comparison is the point, not the figure.

The trap. Leaving the customer discount out. It is a real cost, it is on the order, and it is often the biggest line. A cost-of-sale number that only counts commission looks great and is fiction.

3. Incrementality

What it is. The share of attributed revenue that would not have happened without the affiliate. This is the number finance most wants and the one most programs cannot produce cleanly. Be honest about that in the report.

What a small brand can measure. Three practical proxies, in ascending order of effort:

1. New-customer share

   Of attributed orders, how many were first-time customers? Pull it from Shopify's customer order count on each attributed order. A program where most attributed orders are from new customers is clearly bringing people in. One where most are repeat customers who already knew you is closer to a discount channel than an acquisition channel.

2. Code and link source

   Sneakylink orders came through a click, so the shopper was demonstrably sent by the affiliate. Static code orders are weaker evidence, because the code could have arrived via a coupon site. Report the two separately. A high share of link-attributed revenue is a stronger incrementality signal.

3. A timing or geographic holdout

   If you want a real estimate, pause outreach or a campaign for one region or one two-week window, and compare new-customer orders there against the rest. It is crude, it needs enough volume to mean anything, and it is still far better than a guess.

The trap. Claiming 100%. Nobody believes it. Reporting new-customer share and link share, with a sentence saying "we treat these as the lower bound of incremental revenue", is credible. Reporting all attributed revenue as incremental is not.

4. Active partner rate and concentration

What it is. Two numbers about the shape of the program. Active partner rate is the share of approved affiliates who made at least one sale in the period. Concentration is the share of net attributed revenue coming from the top ten percent of partners.

Where it comes from. Both come from the affiliate list in Reveshare sorted by sales in the period. Count approved, count those above zero, divide. Then take the top tenth and sum their revenue.

What healthy looks like. Active partner rate varies enormously by program type. A customer ambassador program with thousands of members typically sees a small minority active in any given month, and that is fine. A curated creator program should see most partners active. Concentration follows a power law almost everywhere: it is typical for the top tenth to produce more than half of revenue.

The trap. Reading a low active rate as failure. It is often a recruitment style, not a problem. The number to watch is the trend: if active rate falls while total approved rises, you are filling the roster with people who never post, and the join page or approval setting needs attention. And if concentration passes the point where losing three partners would halve the program, say so in the report. Finance would rather know.

5. Payout ratio and cash timing

What it is. Commission actually paid out in the period, divided by commission accrued. And separately, the lag between an attributed order and the cash leaving your account.

Where it comes from. Paid payouts are in the payout history. Accrued commission is on the attributed order list. The lag is your payout policy: monthly on prior-month orders means the average order is paid roughly six weeks after it happens.

Why finance cares. Accrued commission is a liability on the books. A growing gap between accrued and paid means a growing liability, either because affiliates are not requesting payouts or because payouts are failing. The lag is also a cash-flow advantage: you collect the sale first and pay the commission after the return window, which is a better working capital position than most paid channels.

The trap. Ignoring reversals. If refunds after payout are being deducted from next balances, the paid figure will be lower than expected in some months. Note it so it does not look like a payout failure.

6. Time to first sale

What it is. The median number of days between an affiliate's approval and their first attributed sale. It is the leading indicator for everything else in the report, because a partner who sells in the first two weeks almost always keeps selling, and one who has not sold in sixty days usually never will.

Where it comes from. Approval date and first sale date per affiliate, both in Reveshare. Take the median across partners approved in the last quarter, not the mean, because a few very slow partners drag the mean out.

What healthy looks like. For customer ambassadors recruited post-checkout, a first sale inside a month is a good sign, since they share right after buying. For creators, it depends on their content calendar, and two to four weeks is typical for the ones who will work out.

The trap. Reporting it as a single number without the share who never sell. "Median 11 days" is misleading if half the cohort has no sale at all. Report both: median for those who sold, and share who have not sold yet.

7. The one-page template

Here is the report as a single table. Each row has the number, the same number for the previous period, and one line of commentary.

Below the table, three sentences: what improved, what got worse, and the one thing you are changing next month. That is the whole report.

8. What to say when the number is bad

A bad month reported honestly builds more trust than a good month reported vaguely. The structure that works is: the number, the cause, the fix, the date you will know if the fix worked.

• Cost of sale rose. Usually a campaign with a larger customer discount, or a tier promotion kicking in. Say which. If it was a deliberate Q4 push, say the incremental revenue it bought.
• Net revenue fell. Check refunds first, then concentration. If one top partner went quiet, that is a retention action, not a program failure.
• Active rate fell. Almost always a recruitment surge that has not activated yet. Report time to first sale for the new cohort next month.
• Incrementality floor fell. New-customer share dropping often means codes have leaked to coupon sites and existing customers are using them. The fix is link attribution and code hygiene, and you can point to it directly.

Conclusion

The affiliate program is a marketing channel with a cost, a return and a cash-flow profile, and it deserves a report that treats it as one. Net revenue after refunds and discounts. Effective cost of sale next to paid CAC. An honest incrementality floor from new-customer share and link attribution. Active rate and concentration to show the shape of the program. Payout ratio and lag so the liability is booked and the working-capital advantage is visible. Time to first sale as the early warning. Put them in one table with last period beside them, add three sentences and one action, and send it on the same day every month. The CFO will read it, and the program will get the budget it earns.

---

Frequently Asked Questions