The room where restaurant owners tell the truth.

Editor's field notes

Why operators stopped treating delivery apps as the whole online strategy.

The commission is the visible wound. The deeper cut is the customer relationship: a regular can order every Thursday while the restaurant remains unable to reach that diner directly.

Conversations with independent operators tend to begin with the same acknowledgment: marketplace delivery has a legitimate job. It can introduce a restaurant to diners outside the dining room, create demand on a quiet night, and provide delivery logistics that would be expensive to build alone. The mistake is not using a marketplace. The mistake is allowing that marketplace to become the restaurant’s entire online strategy.

The first pressure is visible on every statement. Fifteen to thirty percent — before food, labor, or packaging is paid — leaves little room inside an industry already measured in narrow margins. An operator running $40,000 per month in marketplace delivery — the median in OrderBridge’s operator conversations — routinely pays $8,000 to $12,000 per month when the effective commission lands between twenty and thirty percent. That figure typically approaches or exceeds rent. It can also equal a meaningful part of the kitchen payroll, which makes the line item an operating decision rather than a marketing abstraction.

The second pressure is less obvious and often more durable. The marketplace keeps the diner’s contact details; each repeat order carries an acquisition cost even when the customer already knows the restaurant. A regular who orders the same meal every week can remain invisible to the operator. The platform can market competing restaurants to that diner, while the restaurant that prepared every order cannot send a service update, recognize a pattern, or invite a direct return.

Direct ordering addresses that ownership problem. The software is not the strategy by itself. The strategic change is that an order can become the beginning of a restaurant-controlled relationship instead of another isolated transaction. Each direct order can strengthen a permission-based customer list, improve service recovery, and make the economics of the next order better than the economics of the last one.

The front of house is a separate problem.

Delivery margin and service execution often get collapsed into one technology decision. They are distinct. A point-of-sale system must move modifiers, courses, payments, and kitchen communication through a rush without forcing the operation to work around the software. A direct-ordering system must make pickup and delivery easy while preserving the customer relationship. Neither job disappears because the other one is handled well.

Operators describe the cost of confusing those jobs in operational terms: duplicate tickets, manual re-entry, throttling that arrives too late, menu changes that fail to propagate, and staff watching several tablets while the dining room fills. The strongest setup is rarely the one with the longest feature list. It is the one that assigns each operational problem to a tool built to solve it, then keeps the handoff simple enough for a busy service.

Two problems, two tools.

Operators in the community consistently return to two tools. Toast addresses floor and kitchen — modifiers, courses, payments, and kitchen communication. ChowNow addresses the direct online channel. They complement rather than compete, which is why both recur in conversations.

That distinction matters because a POS migration and a channel shift carry different risks. Replacing the service system touches training, hardware, menus, printers, payment flow, and the kitchen line. Building a direct-order channel touches customer habits, ordering links, menu economics, pickup expectations, and delivery coverage. Operators can sequence those changes instead of treating them as one dramatic switch.

The transition math

Operators who shift a portion of marketplace repeat orders to direct channels typically report the same trajectory. A restaurant moving from a 15% direct / 85% third-party split to a 55% / 45% split, on $40,000 in monthly online revenue, redirects $16,000 from marketplace volume to direct volume. At a fifteen-to-thirty-percent marketplace commission, the gross commission retained is $2,400 to $4,800 per month before direct-platform and payment-processing costs. At the twenty-to-twenty-five-percent rates that recur in many statements, the range is $3,200 to $4,000.

Those figures are arithmetic, not a promise. Menu mix, promotional charges, processing fees, delivery subsidies, refunds, and the direct platform’s fixed cost all change the final result. The useful pattern is still consistent: moving repeat demand changes the margin without requiring the restaurant to invent the demand again. The same $40,000 in online revenue can contribute more to payroll, rent, maintenance, and cash reserves when fewer repeat orders carry a marketplace acquisition charge.

The mechanics vary — QR codes on takeout bags, ordering links moved to the top of Instagram, direct-order incentives in receipt emails. What consistently works is redirecting existing repeat customers, not acquiring new ones. Operators also report that abrupt marketplace exits create unnecessary risk. A measured transition preserves discovery while giving regulars a clear direct path, allowing the mix to change as customer behavior changes.

What this framing is not

OrderBridge is not a software vendor and is not a public review site. Featured tools appear because they recur in operator conversations, not because of commercial arrangements. The community’s disclosure page details the affiliate relationships that exist and how they are structured. Editorial independence is real: no vendor sees this content before publication.

The field notes are also not a universal prescription. Restaurants differ by service model, geography, menu, staffing, and order density. The Editors publish the recurring patterns, the underlying math, and the tradeoffs so operators can examine the assumptions. Verified private threads provide the room for details that should never be attached to a restaurant’s public identity.