The number you’re actually looking at
When I glance at my dashboard first thing in the morning, the number that jumps out is usually gross revenue for the day. It feels good. It’s also not the number that pays my bills, because it doesn’t account for the refunds that will land two days from now or the chargeback that’s sitting in a dispute queue I haven’t checked in a week.
Refunds and chargebacks are the quiet correction to every optimistic number you look at. They’re not exciting to track, which is exactly why so many solo operators don’t bother, and why the ones who do end up with a clearer picture of what’s actually happening in their business.
What a refund actually is
A refund is you giving money back. A customer emails you, or clicks a button in your billing portal, and you (or your payment processor) return the funds. It’s initiated by you, it’s usually quick, and it costs you the transaction amount, sometimes minus a processing fee your payment processor doesn’t return to you even though the money went back to the customer.
Refunds happen for reasons you can mostly understand if you go looking: the product wasn’t what someone expected, they signed up twice by accident, they asked for a trial extension and you said no so they asked for their money back instead, or they just didn’t use it and felt bad charging themselves for something they weren’t touching.
What a chargeback actually is
A chargeback is different, and it’s worth understanding the mechanics because they change how you should think about the number. A chargeback is initiated by the customer’s bank, not by you. The customer contacts their card issuer and disputes the charge, the bank pulls the funds back from your account, and you’re notified after the fact. You often get a chance to contest it with evidence, but the money is already gone while that plays out.
Chargebacks almost always come with a fee from your payment processor, on top of losing the sale itself. That fee exists whether you win the dispute or not, in most processor agreements. And processors track your chargeback rate as a signal of how risky it is to keep working with you. If that rate climbs, you can end up in higher-scrutiny territory with your processor, which is a separate problem from the lost revenue itself.
The practical difference for you as an operator: a refund is a conversation you can have. A chargeback is a decision that’s already been made somewhere else, and you’re reacting to it.
Why these numbers get buried
Most revenue dashboards default to showing gross figures because gross is simpler to compute and looks better. If you’re checking MRR in your billing tool’s summary view, there’s a real chance you’re looking at gross MRR unless you’ve specifically gone looking for a net figure that subtracts refunds and failed payments.
This matters more the smaller your business is. If you’re doing a modest amount of monthly revenue, a single chargeback or a cluster of refunds in one week can swing your growth number from “up” to “flat” or “down.” At a bigger company, that same absolute dollar amount barely moves the needle. As a solo operator, it’s a meaningful chunk of the picture, and if you’re only looking at the topline you’ll misread a month.
Building a refund and chargeback rate
The number worth tracking is a rate, not just a raw count, because raw counts don’t tell you if things are getting better or worse relative to how much you’re actually selling.
Refund rate is refund dollars (or refund count, depending on what you’re trying to see) divided by the revenue or orders for the same period. If you took in $4,000 in a month and refunded $200 of it, your refund rate for that month is 5%. That’s a number you can watch over time and compare month to month.
Chargeback rate works the same way, but it’s usually more useful as a rate of transaction count rather than dollars, because that’s closer to how payment processors themselves calculate it. If you processed 150 transactions and had 2 chargebacks, that’s roughly 1.3% of transactions.
I keep these in a simple spreadsheet next to my regular revenue tracking, not because it needs to be fancy, but because the trend over several months tells me more than any single month does. One bad month with a spike in refunds could be one unhappy customer who filed disputes on several past charges at once. Six months of a slowly climbing rate is a different kind of signal, and it’s worth stopping to ask what changed around the time the trend started.
What the trend can tell you, and what it can’t
A rising refund rate is a prompt to go look, not a verdict on its own. It might mean your product description oversold what the product does. It might mean a change you made to onboarding confused people. It might mean you changed a price and the new price attracted a different, less-committed kind of buyer. It might also mean nothing systemic at all, just a run of unrelated one-off situations.
The metric tells you something moved. It doesn’t tell you why. That part still requires actually reading refund reasons if your platform captures them, or emailing a few of the people who asked for their money back and asking what happened. I’ve had refund clusters that turned out to be a single confused customer who bought the same thing three times, and refund clusters that turned out to be a genuine problem with how a feature was described on my sales page. The rate alone can’t distinguish between those two stories. Only looking closer can.
The same caution applies to chargebacks. A chargeback doesn’t necessarily mean fraud or a genuinely dissatisfied customer. Sometimes it means someone didn’t recognize the charge on their statement and went straight to their bank instead of emailing you first, which is a real behavior pattern worth knowing about even though it has nothing to do with product quality.
Where this fits with the rest of your metrics
If you’re tracking MRR, refunds and failed or reversed payments are part of what separates gross MRR from net MRR. Net is the more honest number for understanding your actual trajectory, because it’s what’s left after the corrections. If you’re tracking churn, a chargeback that results in a canceled subscription should show up in your churn count the same way a normal cancellation would, otherwise your churn number is quietly undercounting what happened.
None of this needs an elaborate system to start. A monthly review where you pull your total refunds, total chargebacks, and the transaction counts behind them, and sit them next to your revenue for the same period, is enough to catch a trend before it becomes a pattern you can’t explain in hindsight.
Keeping it honest
I’ll say plainly what this kind of tracking can’t do. It can’t tell you whether a customer was right to ask for a refund. It can’t tell you whether a chargeback was legitimate or a mistake on the customer’s part. It can’t predict which of next month’s sales will come back as a refund. What it does is give you an honest number to check your gut against, so you’re not running your business off the version of your revenue that only counts the money coming in and ignores the money going back out.
If you want to go deeper on how refunds and chargebacks connect to churn, MRR, and the rest of the numbers you’re already tracking, there’s more on the site.
Find more explainers like this one on the Data Research Analysis Collection home page.
Get new guides and videos first — join the Telegram channel.