The numbers come first. Third-party delivery marketplaces charge restaurants commissions of up to 30% per order, according to Noah Glass, founder and CEO of Olo, the restaurant commerce platform that powers online ordering for more than 800 restaurant brands across 90,000-plus locations, including Shake Shack, Waffle House, Cracker Barrel, Five Guys and Panda Express.
To cover those fees, 82% of brands mark up menu prices on the marketplace apps, Glass writes, and more than half of those markups run between 20% and 30%. The cost lands on the guest at checkout, even as the restaurant loses ownership of the transaction: every order placed through a marketplace belongs to a customer the restaurant never actually owns and must pay, again, to reach in the future.
The alternative, a restaurant's own app or website, solves the ownership problem but creates a scale one. Acquiring a new guest through a first-party app can cost a restaurant up to $100, Glass notes, and most downloaded apps go unopened after the first use.
Glass, who built the first text-message restaurant order in 2005, two years before the iPhone launched, argues the industry has been boxed into a false binary for 20 years: rent reach from a third-party platform or build ownership alone. He is now pushing a third model he calls 'second party' ordering, a shared infrastructure that would let restaurant brands serve guests directly, with saved checkout data and personalized offers, without commissions or markups. He points to Shopify's Shop App, which already lets independent retailers share infrastructure without surrendering their own customer relationships, as the closest existing analog.
Under that structure, Glass writes, restaurants would get marketplace-level reach without giving up either the commission or the guest relationship, while diners would get an ordering experience that reflects their actual preferences instead of generic, one-size-fits-all promotions.
The commentary is Glass's own view and was published as an opinion piece; it does not reflect the editorial position of the outlet that ran it.
The pattern here is a familiar one for anyone who watches how middlemen behave once they capture a market. Delivery platforms became indispensable during the pandemic, then kept extracting rent long after the emergency that justified their rise had passed. The 30% commission and the resulting menu markups function, in practice, as a tax on the consumer, imposed not by a legislature but by a handful of platforms with leverage over restaurants that had no better option at the time.
What is notable is the fix on offer: not a regulatory cap on commissions, not a government mandate, but a private-sector alternative built by entrepreneurs competing for restaurants' business. Capital rewards clear rules and clean ownership of data, and if Glass's second-party model works, it will be because restaurants and guests chose it in the market, not because Washington forced platforms to lower their take. That is the free-enterprise answer to a free-enterprise problem, and it is worth watching whether the marketplaces respond with lower fees of their own once a real competitor for restaurants' loyalty finally exists.
