The Platform Tax: What 90 Minutes With Olo, DeliverThat, Catering Rewards, and RevGen Revealed About Third-Party Catering

On July 15, 2026, New Catering Connections convened a 90-minute open-forum webinar under a deliberately provocative banner: Catering Independence Day. Around 140 operators registered and roughly 90+ attended live to hear five practitioners argue a single thesis — that the third-party catering marketplace, for all its convenience, has quietly become the most expensive line item in the restaurant catering channel.

The panel was assembled to cover the full stack rather than a single vendor's point of view. Byron Duncan, founder of New Catering Connections, hosted alongside Danielle Guzzetta of RevGen Marketing sharing her marketing and program-structure lens. Hillary Holmes, Senior Manager of Partnerships at Olo, spoke to ordering technology. Preet Saini, co-founder of Catering Rewards, spoke as both a software founder and a working restaurant operator. Aaron Hoffman, CEO of DeliverThat, covered the last mile.

Duncan framed the intent of the event in plainly symbolic terms:

This article documents what the panel actually established — the economics, the objections, and the sequencing — because the numbers are more specific and more useful than the rhetoric.

The distinction matters because most operators evaluate marketplaces on the commission alone. The Catering Independence Day framing insists on the fuller accounting. A marketplace order arrives with a fee attached and departs with the customer record. The restaurant produced the food, packaged it, absorbed the labor, and carried the reputational risk of the event going well — yet cannot email that customer next quarter, cannot see what they ordered, and cannot tell whether they came back.

Duncan's core argument was that this arrangement means the operator is, functionally, running a catering program for somebody else's benefit.

Preet Saini offered the most credible testimony because it was his own P&L. He began with roughly 75% of his catering volume flowing through third-party platforms. Top-line revenue looked healthy. The bottom line did not survive the stack of royalties, marketing fees, and platform commissions layered on top of one another. He responded by building his own software and shifting his mix to approximately 40–45% third-party against 60% first-party.

Aaron Hoffman then reframed the question in a way that made the comparison unavoidable. He asked the panel what a typical in-house catering sales representative costs as a percentage of revenue:

Hillary Holmes answered from her previous operating role running a beer hall: catering and event sales reps were paid a 6% commission.

That single exchange produced the sharpest number of the session. A human being whose entire job is to find catering business, build the relationship, and keep the account costs roughly 6%. A marketplace that finds the order once and keeps the customer costs 15–40%.

Hoffman illustrated the consequence with a specific operator:

A 20-unit chain owner in Ohio received a $400 catering order and netted approximately $200 after fees. He turned the platform off immediately, having concluded it was counterproductive.

Hoffman added a second cost that operators frequently overlook: delivery pricing visible to the buyer. When a marketplace attaches $100 to $200 of delivery cost to a catering order, the buyer may simply abandon the purchase. A direct relationship with a catering-specialized delivery provider often reduces that number enough to convert the order that would otherwise have been lost.

The panel's most counterintuitive slide made the case that a restaurant can earn more on $7,500 of first-party catering volume than on $10,000 of third-party volume, because the commission structure consumes the difference and then some. Lower gross, higher net.

The follow-on benefits compound the arithmetic. Panelists put the margin reclaimed by moving an order from third-party to first-party at 15–30%, and the increase in customer lifetime value at three to five times, on the reasoning that a customer you can identify is a customer you can market to again. Data ownership moves from zero to 100%.

Preet Saini quantified what a retained catering customer is worth. A good corporate catering account orders three to four times per year and stays for three to four years. At an average order value near $400, that is a lifetime value in the neighborhood of $5,000. On a marketplace, that same customer is a series of unconnected $400 transactions belonging to somebody else's database.

Every panelist rejected the idea of switching platforms off abruptly. Danielle Guzzetta was the most explicit that this is a migration, not a decision:

Guzzetta recommended making the transition gradually and shifting marketing investment from a 3rd Party Marketplace toward the restaurant’s first-party sales program.

She laid out a maturity ladder that gives operators a way to benchmark honestly rather than aspirationally.

A separate session in the NCC webinar series placed the long-run ceiling even tighter, arguing that third-party marketplaces should ideally not exceed 30% of a brand's catering business.

Guzzetta also offered the single most transferable tactic of the session, which costs almost nothing to implement. Place a sticker on the inside lid of every third-party catering box offering $25 off the next $250 order when the customer books direct. The marketplace delivered the introduction; the sticker converts it into a relationship. Every box already going out becomes an acquisition channel.

Hillary Holmes added a technology-side version of the same idea. Integration tooling such as Olo Rails can route third-party orders through the restaurant's own system, capturing the customer record for a small flat fee — she cited roughly 40 cents — rather than surrendering it entirely.

Holmes contrasted access to customer information through a 3rd Party Marketplace with the opportunities available through restaurant-website ordering. Her point was to evaluate how each ordering channel supports the restaurant’s customer relationships and marketing.

The panel conceded this readily, which is what gave the rest of the argument its credibility. Hoffman acknowledged directly that marketplaces solve a genuine discovery problem for buyers. An office manager who has never heard of a restaurant can find it, compare it, and order from it in a few minutes. That is real value, and pretending otherwise does operators no favors.

The conclusion the panel drew was not that marketplaces are worthless but that discovery is a beginning, not an arrangement. The restaurant's obligation is to have a strategy that converts marketplace-sourced discovery into a direct, repeat, owned relationship. Without that conversion strategy, the restaurant pays the discovery fee again on every subsequent order for the life of the customer.

Holmes addressed the related objection that operators lack the technology to go direct. Catering-specific platforms now provide the functional requirements operators actually need — tax-exempt status handling, house accounts, prep lists — without a 40% margin consequence. The technology gap, in other words, is largely closed; the strategy gap is what remains.

The Catering Independence Day materials organize the entire argument into six outcomes an operator claims by moving first-party. They function well as an internal audit checklist.

Rather than resolve everything in 90 minutes, the panel identified five pillars, each becoming its own deep-dive session in subsequent months.

Program structure, led by Guzzetta and Duncan, begins with a single non-negotiable: someone must own catering. A program without a dedicated leader is a collection of intentions.

Order technology, led by Holmes, argues for catering-specific platforms priced as a flat fee plus a small percentage rather than a commission on gross.

Delivery management, led by Hoffman, treats the last mile as a specialist discipline. He cited a 400-unit brand that saved $1.5 million in fees in a single year by moving to a managed delivery partner.

Rewards and loyalty, led by Saini, converts one-time orders into standing accounts. He was notably firm that loyalty currency must be genuinely useful to the recipient rather than a closed loop designed only to trap spend:

Marketing and sales, again Duncan with Guzzetta, rests on the least glamorous finding in the entire series: catering does not sell itself. It requires proactive outbound effort, structured and scheduled.

The panel's practical conclusions can be stated compactly.

A restaurant should know its current third-party share as a percentage and compare it against the maturity ladder rather than against a competitor's marketing claims. It should calculate what an in-house or fractional salesperson costs as a percentage of catering revenue and set that against blended marketplace fees. It should treat every marketplace order as a lead requiring conversion, using in-box offers and data-capture integrations to make that conversion routine rather than exceptional. And it should recognize that flat revenue across a year is not stability — it is usually a symptom of platform dependency, because genuine first-party catering follows a pronounced seasonal pattern.

That seasonal pattern, and the operating discipline it demands, is the subject of a separate article in this series.

This article is drawn from the Catering Independence Day webinar (July 15, 2026), presented by New Catering Connections with RevGen Marketing, Olo, Catering Rewards, and DeliverThat, and from the supporting event materials.