Most operators think running a discount is the safe way to boost orders. The math says otherwise. Here is exactly how the wrong promo loses you money, and what to run instead.
"Run a promo" is the most common advice restaurants get when their delivery numbers are down. And it's also one of the fastest ways to destroy your margin without realizing it.
I've seen operators celebrate a "successful" promo week because order count went up 40%, without realizing they lost $800 on the week because every order came with a 25% platform-funded discount where they paid most of it.
Let me walk you through the actual math.
The economics of a typical delivery order
On a $20 order, here's roughly where the money goes:
- Platform commission: 25-30% → ~$5-6 to Uber/DoorDash
- Food cost: 28-33% → ~$6-7 to ingredients
- Packaging: 3-5% → ~$0.60-1
- Labor allocation: 20-25% → ~$4-5
- Remaining margin: $2-4 per order
Your average order nets you $2-4 on a healthy day. That's your working margin.
What a "20% off" promo actually does
Let's say you run "20% off entire menu" for a week. Customer sees $20 cart, pays $16. Here's what changes:
- Revenue: $20 → $16 per order
- Platform commission: stays based on $20 gross (they still take their cut)
- Food + packaging: unchanged
- Labor: unchanged
- Your margin: $2-4 → −$1 to $1 per order
So if your order volume increases 30% during the promo week, but your margin per order drops 80%, the math says you're worse off.
What actually works instead
1. Platform-funded promos
Uber and DoorDash periodically run campaigns where they subsidize the discount, meaning the customer sees a lower price but you get paid your full amount. These are always worth participating in. They're the closest thing to a free lunch in this business.
2. Bundle-based upsells
A "combo" that includes an item customers wouldn't normally add (a drink, a side) grows AOV without discounting. The customer feels like they got a deal. You actually make more per order.
3. High-AOV-threshold promos
"Spend $30, save $5" forces customers to order more than they normally would. The $5 off on a $30 order is much less painful than 20% off on $20.
4. Time-of-day promos
Running a discount only during your slowest hours (usually 2-5pm) fills capacity you were paying for anyway. Your fixed costs are sunk. Any order above food + packaging cost is net positive.
Never run a blanket discount greater than 10% unless the platform is funding at least half of it. Full stop.
The retention trap
Here's the other half of the story. Customers acquired through heavy discounts don't come back at full price. You've trained them that your restaurant is discounted.
Over 12 months, a restaurant that relies on heavy promotion typically shows these patterns:
- High volume during promo weeks, low volume otherwise
- Declining average order value over time
- Customer reviews start complaining when prices "jump" (which is actually just returning to normal)
- The promo becomes permanent because revenue collapses without it
That's how operators end up running 20% off 365 days a year, convinced they can't stop because sales will crash. And technically, they're right, but only because they trained their customer base to never pay full price.
The discipline that works
After 6 years and 4 concepts, here's what I've actually seen drive durable growth:
- First 45 days: hit the platform with meaningful promos to build momentum and reviews
- After the honeymoon: switch to bundle logic, platform-funded promos, and AOV threshold offers
- Reserve blanket discounts for launch weeks, slow seasons, and strategic competitive moments only
- Track margin per order, not just order count, the only number that really matters
Boring? Maybe. But it's what actually builds a restaurant that makes money instead of chasing volume at break-even.