The 30% you owe DoorDash, ranked by who's actually losing it

Most restaurants we onboard pay 25–30% in commission to third-party delivery platforms — DoorDash, Uber Eats, Grubhub. The number isn't a secret. The headline outrage about it isn't useful either.

The interesting question is more specific: which line items on your P&L are actually absorbing that commission?

We've reviewed close to 80 restaurant operator P&Ls in the last three years. The pattern is consistent — and the loss almost never lands where the operator thinks.

The intuitive (wrong) answer

Most operators describe third-party delivery as "lower margin but additive volume." The mental model is: net it out, the commission comes off the top, you still made money on incremental orders you wouldn't otherwise have served.

For roughly two-thirds of operators we've audited, this is materially false. Here's the actual decomposition.

Where the 30% comes from, ranked

1. Wages, ~9 percentage points. Third-party orders almost universally take longer to fulfill per dollar than dine-in or direct orders. Higher modification rate. More re-printing. Front-of-house labor pulled into pickup-bag traffic. We routinely measure 25-40% higher labor cost per third-party order.

2. Repeat customer dilution, ~7 points. Customers acquired through DoorDash do not become your customers. They become DoorDash's. We can show this in the data: cohorts acquired through third-party platforms have repeat-purchase rates of 12-22% over 6 months. Cohorts acquired through your own domain — same restaurant, same menu, same prices — repeat at 38-55%.

3. Menu engineering distortion, ~5 points. Operators reflexively raise menu prices on third-party platforms to cover commission. This compresses your premium menu items relative to value items, because customers price-anchor on the third-party menu. Your margin mix drifts toward lower-margin items. This is invisible until you measure it.

4. Brand ambiguity, ~4 points. Your brand becomes one of 200 logos in a list. "I had X for dinner" becomes "I had DoorDash for dinner." Word-of-mouth referral attribution craters.

5. Marketing spend leakage, ~3 points. Operators on third-party platforms keep paying for local discovery anyway — Google, Facebook, OOH — and watch a meaningful percentage of that traffic land on a third-party platform's order page instead of theirs.

6. Pricing transparency loss, ~2 points. You can no longer A/B test pricing, track item-level conversion, or run targeted promotions to specific customer cohorts. The instrumentation isn't yours.

That's the 30%, decomposed. Three of those six items don't appear anywhere on your P&L line by line. They show up as smaller P&L numbers — lower repeat customer revenue, narrower margin on your top items, weaker direct traffic — that you'll probably attribute to "the market."

The on-domain alternative isn't free either

We're not going to claim direct on-domain ordering is zero-cost. Real numbers, from our restaurant client cohort:

• Embedded ordering on your domain: 0.5–2% payment processing, plus the platform cost.
• Customer acquisition via direct discovery (organic + paid): variable, but our average new-customer CAC for direct ordering is roughly $3.40 vs. an effective $15-22 for third-party (when you include the dilution effects above).
• Repeat purchase rate: roughly 2.5x higher.

Net: we have not yet onboarded a restaurant client where the year-one math on shifting 60-80% of off-premise volume to direct didn't work decisively.

What to do this week

If you're a restaurant operator and you've never decomposed your third-party economics this way, do it. You don't need software to do it — you need an honest hour with your P&L, your DoorDash dashboard, and someone willing to argue with the assumptions.

If you'd like an external read on it, book a call — we'll do the decomposition with you, free, in 30 minutes. Restaurants are roughly 30% of our client base. We've done this exercise more times than is probably healthy.