How to Move Delivery-App Customers to Direct Ordering
Here's the uncomfortable part: every customer who orders through DoorDash or Uber Eats is their customer first, not yours. You cooked the food, but the app owns the phone number, the order history, and the "reorder" button. Win them onto WhatsApp or your own ordering link and that relationship — and the 15-30% you're paying to rent it — finally becomes yours.
This isn't about quitting delivery apps. They're still the best discovery tool a new or unknown kitchen has — plenty of first-time customers find you there and nowhere else. The move is narrower and more achievable: take the people who've already ordered from you once, and give them a specific, low-friction reason to order direct the second time. That's the entire playbook below.
Why moving delivery-app customers to direct ordering pays off
Third-party commission runs 15-30% on DoorDash, 15-30% on Uber Eats, and around 20-25% on Grubhub, depending on plan tier (Rezku, 2026). That's before card processing and any promo costs the platform passes back to you. Move a repeat customer from a 25%-commission order to a direct WhatsApp order and you're not finding new margin, you're just keeping margin you already earned. Our commission maths breakdown walks through the exact pounds-and-pence version for a typical week; if you haven't run those numbers for your own menu, do that first — it's what makes the case for spending any effort on this at all.
You're also not alone in trying. A 2026 Market Leader survey of roughly 450 restaurant operators, run by Restaurant Business and Nation's Restaurant News, found 53% are actively working to reduce their reliance on third-party delivery, and a growing share are doing it by tying loyalty perks specifically to direct orders rather than dropping the apps outright (Restaurant Business, 2026). That's the sane middle path: keep the apps for discovery, build the direct channel for retention.
The ask has to happen off the app
You can't message a DoorDash or Uber Eats customer to pitch them on ordering direct next time — you don't have their number, and soliciting them inside the platform isn't the game plan anyway. The ask has to live somewhere the customer sees after the app has done its job: the food itself.
That means three touchpoints, in order of how well they convert:
- A card or sticker inside the bag. Cheap to print, impossible to miss, and it's the one moment you have the customer's full attention — they're about to eat.
- A line printed on the receipt or order sticker, if you don't want a separate insert. Weaker than a standalone card, but free if you're already printing dockets.
- A follow-up message, but only for customers who've already messaged you once directly — a delivery-app customer who calls with a question, or one who leaves a review you reply to. That's a real, permission-based opening; cold-messaging someone from their delivery address isn't.
The insert that actually gets scanned
A QR code on restaurant flyers and inserts scans at roughly 45% — well above retail (around 6%) and event flyers (around 18%), according to a 2026 study of over 950 restaurants (Uniqode, 2026). Food gets people to pull out their phone in a way a shop receipt never does. Use that.
What the card should actually say — keep it to one offer, one action, one code:
Order direct next time, save {X}%. Scan to message us on WhatsApp — no app, no delivery fee, we confirm in minutes. Code:
DIRECT10
Three things make or break this card:
- The QR code has to open a chat, not a menu PDF. Point it straight at your WhatsApp link (
wa.me/yournumber) with a pre-filled message like "Hi, I got this card with my order" — it removes every step between scanning and ordering. - The discount has to beat what the app already gives them. If DoorDash regularly runs 20% off for this customer, a 10% direct-order code loses. Check what promotions are currently live on your own listing before you set the number.
- Print it before the busy season, not during. A card designed the week you need it is a card that goes out with a typo or a dead link. Batch-print a few hundred once your WhatsApp ordering setup is solid enough to handle the volume.
Picking the incentive: discount, freebie, or loyalty credit
All three work; they solve slightly different problems.
| Incentive | Best for | Trade-off |
|---|---|---|
| Percentage discount (e.g. 10-15% off) | First-time direct order | Easiest to explain on a card; costs you margin every time it's redeemed |
| Free add-on (a drink, a side, a dessert) | Menus with a cheap, high-margin item to give away | Feels bigger than its cost to you; harder to fit on a small insert |
| Loyalty credit / punch card | Customers who already order from you more than once a month | Builds a reason to keep ordering direct, not just switch once |
Don't stack all three. Pick one, run it for a full month, and see whether direct orders actually move before adding a second offer. Our WhatsApp loyalty and repeat-customer guide covers how to structure the ongoing version once the initial switch has worked — a punch card or points system that rewards staying direct, not just arriving there.
What to say once they've scanned
The scan gets them into a chat. What happens in the next thirty seconds decides whether they order or close the app. Don't make them figure out your menu from scratch — confirm the offer immediately and make ordering the very next step:
Hey! 👋 Thanks for scanning — here's 10% off your next order when you message us direct (code DIRECT10, no expiry this month). What can we get started for you?
If you're set up with a saved catalogue, send it straight after. If not, ask the two or three questions you'd ask any WhatsApp order — what they want, pickup or delivery, and when. Keep the reply speed as fast as their first DoorDash confirmation was, or the whole pitch — "direct is faster and cheaper" — falls apart on the first try. Our order message templates have the confirmation scripts ready to copy in rather than typing from scratch mid-rush.
Where this doesn't work
Be honest about the limits, because forcing it wastes print runs and goodwill:
- Apartment or office deliveries with no direct line of sight to the customer — if a doorman or reception desk takes the bag, your insert may never reach the person who ordered. Lean harder on the receipt line here, since it travels with the food itself.
- One-off, out-of-area customers who ordered you once because they were visiting or the local place was closed. They're not switching channels for a single transaction; don't spend a premium incentive on them.
- Delivery-only ghost kitchens with no pickup option. Direct WhatsApp ordering still works, but you'll need your own delivery arrangement or a local courier — read our zero-commission ordering guide before promising delivery you can't yet fulfil.
Track it or you won't know if it worked
Give the code or the QR link its own tracking — a distinct WhatsApp link (wa.me/yournumber?text=DIRECT10) or a simple tally when the code gets typed in chat. After four weeks you want one number: how many direct orders came in that quoted the card, against how many bags went out with one inside. A 2-4% redemption rate is a realistic, healthy start for a first run — treat anything above that as a sign to print more, and anything near zero as a sign the offer, not the tactic, needs to change.
The goal was never zero delivery-app orders. It's a slowly growing slice of your regulars who now order the way that keeps more of what they pay you — one bag, one card, one scan at a time.