Gross, Net of Refunds, or Without Rebills: Which Revenue Number to Report

Ask three people in the same company what last month’s revenue was and you can get three honest, different answers. The media buyer says $310,000, counting every charge. Finance says $265,000, net of refunds. The founder, thinking about which campaigns are working, wants to know how much came from new buyers this month, not instalments from customers who signed up in spring. Deciding whether to report gross revenue, revenue net of refunds, or revenue without rebills isn’t pedantry. Each number answers a different question, and using the wrong one for a decision is one of the quieter ways marketing reports mislead.

This is especially sharp for businesses that sell high-ticket offers on calls, where payment plans, refunds and rebills are normal. A $9,000 programme sold on a six-month plan, with a 10% refund rate, produces very different “revenue” depending on which cut you use and when you look. This post defines each number, shows where each one belongs, and offers a reporting setup that keeps everyone honest.

The three numbers, defined

Gross revenue is every charge collected in the period, before anything comes back out. Full-pay purchases, deposits, instalments, renewals, all of it.

Revenue net of refunds is gross revenue minus refunds issued. There’s a subtlety: are you subtracting refunds issued in the period (cash view) or refunds on sales made in the period (cohort view)? Finance usually wants the first. Marketing decisions usually need the second.

Revenue without rebills (first payments only) counts only the first payment from each sale. Instalments two through six, subscription renewals and repeat charges are excluded. It’s a measure of new money generated by this period’s acquisition, not money collected from past acquisition.

There’s a fourth number worth naming: contracted revenue, the full value of what was sold (all six instalments of the $9,000 plan) on the day it was sold. It’s useful for sales compensation and pipeline, but it’s a promise, not cash, and it shouldn’t be confused with any of the three above.

Why the gap between them gets big

Take an illustrative month for a coaching business:

• 40 sales of a $9,000 programme. Half pay in full, half on a six-month plan of $1,500.
• 6 refunds in the month: 3 on full-pay sales from this month, 3 on full-pay sales from last month.
• $120,000 in instalment payments from customers who bought in previous months.

Measure	Calculation	Amount
Contracted revenue	40 × $9,000	$360,000
Gross revenue	20 × $9,000 + 20 × $1,500 + $120,000	$330,000
Net of refunds (cash view)	$330,000 − 6 × $9,000	$276,000
First payments only	20 × $9,000 + 20 × $1,500	$210,000
First payments, net of this month’s refunds	$210,000 − 3 × $9,000	$183,000

Same business, same month. The highest number is nearly double the lowest. If marketing spent $70,000, ROAS ranges from 2.6x to 5.1x depending on which figure goes in the numerator. That’s the difference between “comfortably profitable, scale” and “fine, be careful.”

Which number to use for which decision

For judging campaigns and creative: first payments, ideally net of refunds

When you’re deciding whether an ad, ad set or campaign is working, you want the revenue it generated from people it acquired. Rebills from customers acquired months ago by a different campaign shouldn’t inflate this month’s numbers for whatever ad those customers happened to click first. Revenue without rebills strips that out.

Refunds matter here too, maybe more than anywhere. Some campaigns attract buyers who refund at twice the rate of others. An angle that over-promises can look like your best performer on gross and be your worst on net. Look at refund rate by campaign at least monthly. We dig into this pattern in attribution-adjusted ROAS.

For LTV and payback: all payments, net of refunds, by cohort

Lifetime value is the opposite case. You want every payment a customer ever makes, rebills included, minus anything refunded, grouped by when they were acquired. A campaign whose buyers pay a $1,500 instalment and then default looks very different from one whose buyers complete all six payments, and you only see that with all payments counted. See CAC and LTV explained for the cohort mechanics.

For finance and the P&L: net of refunds, cash view

Finance reports what hit the bank account and what left it in the period. That’s gross collected minus refunds issued, regardless of when the original sale happened. Marketing doesn’t need to report this number, but it does need to reconcile to it. When the CFO asks why marketing’s revenue doesn’t match finance’s, “we use first payments by cohort for campaign decisions” is a good answer. “I’m not sure” is not.

For board and leadership updates: pick one, label it, and stay consistent

Leadership mainly needs trends and consistency. Choose the number that matches the decision they make (usually net revenue for the business overall, plus first-payment revenue and CAC for growth), label it explicitly on every slide, and don’t switch definitions between quarters. For more on this, see how to present marketing results to a skeptical CFO.

A quick decision guide

Question you’re answering	Revenue number
Which ads and campaigns are working?	First payments only, net of refunds on those sales
What’s our true ROAS this month?	First payments, net of refunds, with gross shown alongside
What is a customer worth over time?	All payments, net of refunds, by acquisition cohort
What did we collect this month?	Gross collected
What does the P&L show?	Net of refunds issued (cash view)
How much did sales sell?	Contracted value, separately from cash

The timing problem with refunds

Refunds lag. A refund policy with a 14-day or 30-day window means this month’s “net” revenue will keep falling for weeks after the month closes. Two practical rules:

1. Don’t judge a campaign’s net revenue until its refund window has passed. For a 30-day guarantee, last month’s net numbers are reliable; this week’s aren’t.
2. Track refund rate by campaign over a rolling 90 days, so a single bad week doesn’t swing decisions.

This is also why you should be careful about which conversions you send to ad platforms. If every purchase is sent the moment it happens, the platform optimizes toward buyers who look good on day one, including future refunders. Some teams send conversion value net of expected refunds, or hold back payment types that tend to unwind, like certain financed purchases.

Rebills and payment plans: count sales once

A related trap: counting each instalment as a “sale.” If a buyer on a six-month plan generates six payment records, a naive report shows six sales. Your close rate, cost per sale and average order value are all wrong.

The same trap catches attribution. If each instalment is credited to whatever ad the buyer clicked most recently before that payment, a customer who clicks a retargeting ad in month four can move their fifth instalment onto a campaign that had nothing to do with the original sale. The cleanest rule is that a sale’s revenue, including later instalments, stays with the click that came before the sale itself. Ask your attribution tool how it handles this before you trust revenue-by-campaign on a business with payment plans.

Count sales by buyer, not by payment record. In Callix, the Unique Sales metric counts sales once per prospect, so a buyer who pays in instalments is one sale, and it’s available in widgets, tables and the Ads API.

How to set up reporting so everyone sees the right cut

The cleanest approach is a dashboard where each card states which cut it uses, so nobody confuses one number for another.

In Callix, when you add or edit a dashboard widget that includes a revenue metric, two options control the cut: Refunds lets you choose the original charge (before refunds) or after refunds, and Payments lets you choose all payments or first payments only. After refunds and all payments are the default, and when a card uses a different cut, its title names it, so a “first payments only” ROAS card can’t be mistaken for the default.

A setup that works for most call-driven businesses:

• Growth dashboard: cost per booked call, close rate, first-payment revenue and ROAS by campaign, refund rate by campaign.
• Business dashboard: gross collected, net of refunds, cash from rebills, LTV by acquisition month.
• Ad platform comparison: what Meta and Google report for their own ads, next to revenue from your payment processor. The gap is informative on its own.

A note on currencies and multiple processors

Two more things can distort revenue before you even choose a cut. If you take payments in more than one currency, or your processor settles in a different currency from what customers pay, adding amounts together without converting them produces nonsense. And if revenue comes from Stripe, Whop, a CRM and manual entries, make sure each source is counted once. Callix converts each payment at the rate on the day it was paid before adding it up, but whatever tool you use, check how it handles both before you trust its totals.

The short version

Gross revenue tells you what came in. Net of refunds tells you what you kept. First payments tells you what this period’s marketing actually produced. Use first payments, net of refunds, to judge campaigns. Use all payments, net of refunds, by cohort, for LTV. Reconcile to finance’s cash view. And label every number, every time, because the most common revenue reporting mistake isn’t choosing the wrong cut. It’s not saying which cut you chose.