Reconciling Your Payment Processor Against Your Own Numbers

Why your dashboard and your bank account tell different stories

If you run a subscription business, you probably have two places where “how much money did I make” lives. One is your own system: your database, your billing table, whatever spreadsheet or app tracks subscribers and charges. The other is your payment processor’s dashboard, Stripe or PayPal or whatever you use to actually move money.

Most of the time these two roughly agree. Then one month they don’t, and you’re staring at a number that’s off by a few hundred dollars and you have no idea why. That gap is what payment processor reconciliation is for. It’s the habit of checking your own records against what the processor actually reports, on a regular schedule, so a small discrepancy gets caught while it’s still small and explainable.

This isn’t glamorous work. Nobody starts a business because they love matching numbers in two spreadsheets. But if you’ve ever found out three months late that a webhook silently failed and forty customers got billed without your system knowing about it, you already understand why this matters.

What “your own numbers” actually means

Before you can reconcile anything, you need to be clear on what your internal numbers represent. For most solo operators, that’s whatever your app or billing system logged as a charge attempt or a successful payment. If you’re calculating MRR from your database, that number is a reflection of what your code thinks happened, not necessarily what happened.

The processor’s numbers are different. Stripe or PayPal know what actually hit the rails: successful charges, failed charges, refunds, disputes, and the fees they took out before depositing the rest into your bank account. Your internal system might not know about a failed charge if the webhook that reports it never fired, or never got processed correctly.

So reconciliation isn’t really “check that two numbers match.” It’s “find out which system is telling the truth about what happened, and figure out why the other one disagrees.”

The four things worth comparing

You don’t need to reconcile every field the processor exposes. For a one-person business, four categories cover almost everything that matters.

Gross charges. The total amount your customers were charged in a given period, before fees or refunds. This is the number that should be closest to what your internal system thinks it billed.

Fees. What the processor took for itself. This varies by plan and card type and sometimes by country, so it’s rarely a clean flat percentage. You’re not trying to predict this number exactly, you’re trying to make sure it’s in a range that makes sense given your processor’s published rates.

Refunds and chargebacks. Money that went out after it came in. These should show up in your internal system too, if your refund process actually updates your database when you issue one. A lot of small mismatches trace back to a refund that was issued from the processor’s dashboard directly and never logged anywhere internally.

Net deposits. What actually landed in your bank account. This is gross charges minus fees minus refunds, roughly, though timing can shift things across periods.

If you compare just these four categories once a month, you’ll catch the overwhelming majority of real problems.

Why the numbers will never match exactly, and that’s fine

Here’s something worth saying plainly: a perfect match between your internal numbers and your processor’s numbers is not the goal, and chasing it will waste your time. A few sources of mismatch are completely normal and don’t indicate anything is wrong.

Timing is the biggest one. A charge made on the last day of the month might get deposited in the first few days of the next month. If you’re comparing calendar months, that charge shows up in different periods depending on which system you’re looking at.

Currency conversion is another. If you accept payments in multiple currencies, the exchange rate your processor used at the moment of conversion won’t exactly match whatever rate you might use internally to normalize everything to your reporting currency.

Retried failed payments are a third. A card might get declined, retried automatically two days later, and succeed. Depending on how your internal system logs charge attempts versus successful charges, that can create a small offset that isn’t an error at all.

The goal isn’t zero variance. The goal is a variance you can explain. If you can point to timing, currency, or retries as the reason for a gap, you’re fine. If you can’t explain it, that’s the signal to dig further.

What an unexplained gap is usually telling you

When the numbers don’t match and you can’t chalk it up to timing or currency, it’s usually one of a few things.

A webhook that silently stopped firing is common. If your billing system relies on the processor telling it “this charge succeeded” or “this subscription renewed,” and that notification gets dropped for any reason, your internal system will undercount. This is the scenario that hurts the most because it can run for weeks before anyone notices, since your database looks internally consistent even though it’s missing events.

A double charge or duplicate subscription is another. If a customer’s signup flow has a bug that can create two subscription records for the same person, the processor will show two charges where your reporting logic expects one, or vice versa if only one made it into your database.

Manual actions taken outside your system show up too. If you or a support person issues a refund directly in the processor’s dashboard rather than through whatever internal flow updates your database, your records will be wrong until someone notices and fixes them by hand.

None of these are catastrophic on their own. They’re normal failure modes of running a small operation without a dedicated finance team. The point of reconciling regularly is that you catch them while the fix is still a five-minute database update, not a multi-month cleanup project.

A simple monthly habit, not a full audit

You don’t need accounting software or a bookkeeper to do basic payment processor reconciliation as a solo operator, though at some point involving an actual accountant for your books is a good idea and outside what this article covers. What you do need is a repeatable routine.

Pull the processor’s monthly summary report. Most platforms give you a breakdown of gross charges, fees, refunds, and net deposits for a given period without much digging. Pull the equivalent numbers from your own system for the same period. Line them up side by side. If they’re close, and any gap is explainable by timing or currency, you’re done for the month. If they’re not close, start narrowing down: check for failed webhooks first since that’s the most common culprit, then look for duplicate records, then check for refunds issued outside your normal flow.

If you’re comfortable writing even simple scripts, automating the pull-and-compare step is worth it. A script that fetches both totals and flags anything outside a reasonable tolerance, say a few percent, turns this from a monthly chore into something that can run itself and only bother you when there’s actually something to look at.

What reconciliation can’t do for you

It’s worth being honest about the limits here. Reconciling your processor against your own numbers tells you whether your records are accurate. It doesn’t tell you why a customer churned, whether your pricing is right, or whether a spike in refunds reflects a product problem or a marketing problem that attracted the wrong customers. It’s a data integrity check, not a business diagnosis. Treat it as the foundation your other metrics sit on, not as an answer in itself. If your MRR or churn numbers are wrong because the underlying charge data was wrong, no amount of clever analysis on top will fix that. Get the base numbers right first, then do the analysis.

It also won’t catch everything. A charge that succeeded on the processor’s end but should never have happened, because a coupon logic bug let someone pay $1 instead of $49, will show up as a perfectly reconciled transaction. Reconciliation confirms your systems agree with each other. It doesn’t confirm your business logic was correct in the first place.

Treat it as a habit, not a project. A few minutes a month spent comparing four numbers is cheap insurance against finding out six months from now that your revenue reporting has been quietly wrong the whole time.

If you want more of this kind of plain-English breakdown of the metrics and methods that actually run a small business, you can find the rest of what we’ve covered on the home page.

Get new guides and videos first — join the Telegram channel.