What each channel is actually for
Marketplaces sell reach: they put a restaurant in front of people who were not looking for it, and charge a percentage of every order for the privilege. A direct storefront sells repeat: it costs payment processing and little else, but brings no new customers on its own. The commission is a customer-acquisition fee — worth paying on acquisition, and expensive on everything after it.
Which reframes the decision. The question is not "should I leave the marketplaces" — for most restaurants the answer is no. It is "how many of my regulars am I still paying an acquisition fee for?"
The maths, without the sales pitch
Take a guest who orders once a week for a year at your average delivery ticket. On a marketplace you pay commission fifty-two times. Through your own storefront you pay it zero times, and card processing on all fifty-two.
The gap on that single customer is usually a meaningful multiple of what a direct ordering system costs to run for the year. That is the whole business case, and it does not require leaving any platform.
The honest counterweight: a marketplace order you would not otherwise have had is worth its commission. Restaurants that pull off the platforms to save fees, and lose the discovery with it, usually go backwards. Both things are true at the same time.
Moving regulars across, without gimmicks
- Put a card in every bag. The most effective tactic and the least used. It reaches somebody who has already decided they like your food.
- Make the direct channel genuinely better, not merely cheaper — a wider menu, items the platforms do not carry, earlier pickup slots, order-ahead for the lunch rush.
- Own the reorder. A one-tap repeat of a previous order beats any discount.
- Do not disparage the platforms on your own packaging. It reads badly, and it breaches most partner agreements.
- Read your agreements before discounting. Several marketplace contracts contain price-parity terms; know what yours says before advertising a cheaper direct price.
The menu-pricing question everyone gets wrong once
Should marketplace prices be higher than in-house prices, to cover the commission?
Most operators eventually say yes, and most arrive there after a year of quietly absorbing the fee and wondering where the margin went. The arguments against are that guests do sometimes notice, and that some agreements restrict it.
What matters mechanically is whether your system can hold per-channel prices on one menu. If marketplace pricing means maintaining a second menu inside each platform portal, you will drift — and the drift always favours the platform, because the edit you forget is the price rise you never applied.
So ask any vendor: can I set a per-channel price on the same item, from one place, and does an 86 reach every channel at once?
What to get right operationally
- One kitchen screen. A tablet per platform means somebody retypes orders during the rush, and retyping is where the refunds come from.
- One menu, one 86. Running out on one channel has to mean running out on all of them.
- Pause, do not throttle. When the kitchen is buried, being able to pause a channel for twenty minutes is worth more than any prep-time setting.
- Auto-pause when you go offline. Accepting orders nobody can see is the worst available failure.
- Reconcile the payouts. Marketplace settlements arrive net of commission, adjustments and refunds. If nobody matches them back to orders, errors are invisible.
Where ThaliPOS fits: the direct storefront and QR codes carry no commission; marketplace channels run on the same menu with per-channel pricing, a fast 86 path, per-channel pause and mandatory auto-pause when the store is unreachable. The marketplace integrations are in partner certification. The ThaliPOS online ordering page →