The deposit hits your account on a Tuesday. One lump sum from DoorDash, and it's a little lighter than the week felt. You've got a lunch rush starting in twenty minutes, so you glance at the number, decide it's probably right, and get back to the line.
Almost every owner does the same thing, because the deposit never lines up with your sales anyway. By the time it lands, the app has skimmed commission, promo splits, refunds, and the occasional chargeback right off the top. Tracing that back to actual orders feels like a lost afternoon, so you trust the number and move on.
Here's the catch. The errors in those payout statements tend to run in one direction, and it isn't the one that helps you. A missing order here, a commission rate that quietly ticked up there, a refund you never approved. None of it looks like much on a single Friday night. Stack it across a year and it's a real line on your books.
Why the deposit never matches your sales
A third-party payout is a net number. It's your gross delivery sales minus a stack of deductions, and the apps don't exactly lay that stack out for you in plain sight.
Walk through what gets pulled before the money reaches you:
- Commission, usually 15 to 30 percent depending on your plan, taken on every order.
- Promo and marketing costs, where you cover part of a discount or an ad placement you opted into (or sometimes one you forgot you opted into).
- Refunds and adjustments, when a customer said the order was cold or a side was missing, and the app refunded them from your pocket.
- Chargebacks and fees, the catch-all for disputed charges and processing costs.
Each one is defensible on its own. The problem is you're seeing the result, not the math, and the report that shows the math lives three screens deep in a dashboard you open once a month, if that.
What a reconciliation automation actually does
Reconciliation is just a plain word for matching two lists and finding what doesn't agree. On one side, every delivery order your point-of-sale system rang up. On the other, every order the app says it paid you for, with its deductions.
An automation does the boring part. Once a week it pulls the payout report from each delivery platform, pulls the matching sales from your POS, and lines them up order by order. Then it hands you a short list: orders that show on one side but not the other, fees that changed from last week, refunds above a dollar amount you set.
You're not reading a spreadsheet. You're reading a handful of flags, each one a question worth asking. Figure ten minutes on a slow afternoon instead of a Sunday at the kitchen table.
Setting it up without a finance degree
You don't need to build this from scratch, and you don't need to understand the plumbing. Here's the shape of it.
- Pick your sources. List the delivery apps you actually use and the POS you ring orders on (Toast, Square, Clover, whatever it is). That's the whole input.
- Get the reports flowing. Each app can email or export a weekly payout report. The automation grabs that file on its own, so nobody has to remember to download it.
- Match on order number and total. The system pairs each POS order to its payout line using the order ID and the amount. Most match cleanly. The ones that don't are your list.
- Set your thresholds. Tell it what you care about: a refund over five dollars, a commission rate that moved, any order on the POS with no payout at all.
- Put it where you'll see it. A short weekly email or a text. Not another login, not another tab you'll forget.
Standing it up is a few hours, most of it spent confirming logins and deciding your thresholds. After that it runs quietly, and you only hear from it when something's worth a look.
What to question, and what to let go
Not every gap is a fight worth having. Part of a good setup is knowing which flags to chase and which to shrug off.
Worth questioning
- An order on your POS with no payout. The kitchen made the food, the money never came. That's the clearest one.
- A commission rate that jumped without you changing plans. Rates do change, but you should know when and why.
- Refunds you never saw. If the app refunded a customer and charged you, you're owed the reason. "Customer said it was late" on an order that went out on time is disputable.
- Duplicate fees or charges that don't tie back to any order.
Usually not worth it
- A few cents of rounding between systems. That's normal, and chasing it costs more than you'd recover.
- A promo you did opt into, even if the split stings. That's the contract, not an error.
The point isn't to dispute everything. It's to walk into the week knowing the three things that are actually off, instead of suspecting everything and proving nothing.
Where the tool stops and you start
Here's the honest part. This automation will not fight the dispute for you, and anybody who tells you it does is selling something.
It can't log a claim with DoorDash support, argue a refund, or sit on hold. Those portals are slow by design, and a human (you or someone on your crew) still has to file the claim and follow up. No tool changes that.
What it changes is the part that actually eats your time: finding the problems. Instead of scrolling statements hoping something jumps out, you get the specific orders, the specific amounts, and the reason each one's flagged. Filing might take ten minutes a week. Deciding what to file used to take the whole afternoon.
One more limit: it can't catch what your POS never recorded. If an order didn't get rung up right, there's nothing to match it against, so clean ticketing still matters.
Worth doing this week
You don't have to automate anything to start. A few small moves this week tell you whether there's money sitting in those statements.
- Open last week's payout report from one app and find the commission line. Just know your real rate.
- Pick five delivery orders and trace them from your POS to the payout. See if they match.
- Write down your thresholds: the refund amount and the fee change that would make you look twice.
- Count the minutes it took you. That's the afternoon you'd get back every week.
- If the gaps turn out to be real, that's when an automation earns its keep. We set these up for Rhode Island shops, and we're a ten-minute drive from most of them.