Teardown · Food-delivery apps
The reorder flow in food-delivery apps
Last reviewed
- Jobs-to-be-done
- Funnel analysis
- RICE
Context
Most major food-delivery apps now offer a way to repeat a past order in one or two taps. It looks like a small convenience. It is better understood as a strategic bet: much of food ordering is habitual, and every screen between a hungry person and the checkout is a chance to lose them to a competitor or to the fridge.
User and job-to-be-done
Primary user: a returning customer who orders from a small set of familiar restaurants.
Job: "When I'm tired and hungry, help me get the meal I know I like, with as little thinking as possible."
Note what the job is not: discovery. The reorder user has already decided. Anything that reopens the decision — banners, new restaurant suggestions, upsell prompts — is friction for this job, however useful it may be for others.
Flow walkthrough
A typical reorder path observed across apps:
- Open app → home feed with promotions and categories.
- Find "order again" or "past orders" — sometimes on the home screen, sometimes in the profile tab.
- Tap a past order → cart pre-filled.
- Handle availability: an item may be unavailable or the restaurant closed.
- Confirm address, payment and any changed delivery fee.
- Place order.
What works
- Pre-filled carts remove the most cognitively expensive step: reconstructing the order.
- Visual recognition: showing dish names and the restaurant together makes the right order easy to spot.
- Saved address and payment make step 5 near-instant for most returning users.
Friction and hypotheses
| Friction | Hypothesis |
|---|---|
| Reorder entry point below the fold or inside the profile | Returning users who scroll past promotions are likelier to abandon or get distracted into browsing |
| Unavailable items silently removed | Users who notice late may lose trust in the cart and re-check everything |
| Price or fee changes surfaced only at checkout | Surprise at the last step may push some users to abandon |
| No "usual" modifications remembered | Users re-enter the same customisations every time |
Metrics we'd watch
- Reorder adoption: share of returning-user orders placed through reorder.
- Time to order for reorder vs browse paths.
- Reorder cart abandonment, segmented by whether an item was unavailable.
- Guardrail: discovery-driven orders and new-restaurant trial, to ensure we are not starving the marketplace.
If we were the PM
- Promote reorder for habitual users only. Show a prominent "your usual" card when recent history suggests a habit. Trade-off: less space for promotions on the home screen.
- Handle unavailability up front. When a past item is unavailable, say so on the reorder card and offer the closest substitute. Trade-off: substitution logic needs restaurant-level data quality.
- Explain price changes before the cart. Show "₹ more than last time, because…" on the card. Trade-off: transparency may reduce some orders in the short term while building trust.
We would test proposal 1 first: it has the broadest reach and the cleanest measurement, with discovery-order share as the guardrail.