Skip to content
Manan Agarwal

PRD — Self Pickup from restaurants on Swiggy

PRD, Product2 min read

This is a fictional prd that I wrote as part of my assignment for a PM internship at Bounce. All metrics are guesstimates. This was written based on a template that can be found here.

For the unitiated, a PRD is a Product Requirement document, a document that tells you what exactly you are building. It outlines the product's purpose, its features, functionalities, and behaviour.

This is my first attempt at writing a PRD so please feel free to dm me any suggestions/feedback on Twitter.

Problem Statement:

Swiggy is evaluating if it should start pickup from restaurants as a new feature on it’s app.

You are the PM on the project, what do you do next?

  • Please list down assumptions you have made

  • How would you plan to launch the project?

  • How will you evaluate success/failure of the launch

Self Pickup from Restaurants

This spec discusses about (WHY), (WHAT), (WHEN) needs to be built for initiating self pickup from restaurants. It does not discuss the technical implementation (HOW) for it.

Author:Manan Agarwal, Fictional PM of Swiggy

Last updated: November 10th, 2020

Why should we build it?

  • Competition parity

  • Engagement/Retention metrics

  • OKRs (Vision: build better urban lifestyle)

  • After the pandemic people are apprehensive of using delivery partners because of their multiple interaction with various elements in the system.

Problem it solves

  1. Wait in restaurant queues to place your order
  2. Wait additionally for the order to be made
  3. High delivery costs incurred during peak hours when ordering from near your home
  4. Swiggy’s end reduces burn wrt rider costs


  • Growth in gross volume of orders.

  • Increase in the retention, DAU and corresponding lagging metrics eg: less time spend for in app ordering

  • Major UX changes, on the customer side. (different swiggy pickup, like in blr airport)

  • Consumers willing to trade picking orders to save time.

Main User Stories

  • As an office goer, I may pickup snacks/coffee/lunch/dinner based on time I go to office, thus reducing my total time elapsed to order and get food.
  • As a home-Maker, I’ll be able to bring food home on a regular basis for the family, since I do it almost everyday.
  • As a college student, I don’t want to spend money on delivery charges and with a packed schedule I cannot wait for my order to get ready
  • As a restaurant owner, self ordering kiosks are very expensive and due to Covid I wouldn’t want multiple people using the same kiosk and hence your phone becomes a kiosk.

Defining the goal of the feature:

Offering unparalleled convenience to urban consumers by eliminating the last mile time and introducing self pickup from restaurants allowing consumers to have maximum flexibility with respect to their order.

Framework Used: DHM (Delight, Hard to Copy, Margin Enhancement)

Potential metrics to watch out for / impact

  • Orders done daily

    • For Bangalore, approximately 60% of all orders in a day form category 1, 10-15% form category 2, additional 15% would form category 3 of people using our app \
  • Orders within distance

    • Understanding from data how many orders have been placed within 1-2km radius of the consumer. A fair estimate would be that ~10% of all orders would fall in the under 1.5km range and ~25% would be under 500meters from restaurants.

Success metrics

  • On restaurant end:
    • Additional GMV generated by pick up(5-10% additional)
    • Completed orders per week in pickup mode (>24)
    • Percentage of orders declined both normally + in peak hour (<5%+<2%)
    • Waiting Minutes (<2mins)
    • Decline in total # of delivery through delivery partners
  • On customer end:
    • Average basket size(>Rs 380)
    • Increase in DAU (7%)
    • Weekly orders (>2)

Product Requirements (The What)

Homepage to have a specific button which lists “Pickup your order”MUST HAVE
Select the restaurant you wish to pick up your food fromMUST HAVE
Show the menuMUST HAVE
Place the order, adding a cart featureMUST HAVE
Create payment flowMUST HAVE
Provided with an ETAMUST HAVE
Alert user via email about order being placedMUST HAVE
Alert via push notif, order is placed and ETA screen should pop upMUST HAVE
Alert via push notif, when food is ready to be picked upMUST HAVE
Alert via push notif, when order is delivered and give prompt for feedback in terms of 1-5 rating overall experience with a comment box.MUST HAVE
Transaction is secured by a One time pin shared on both merchant and consumer end to authenticate during the pickup.SHOULD HAVE
If customers choose to opt-in to share their location data when the app is running, merchants will be alerted when the customer is approaching the store so they can have their order ready upon arrival.SHOULD HAVE
SMS ETA < 2mins, if the customer loses connectivity for the time being.Delighter (Should have)
Call the customer, if the pickup has been waiting in the restaurant for >2mins, to reduce the wait time for the queue.Delighter (Should have)

Non-Product Requirements

Marketing: Update Swiggy Blog on the new featureMust HaveMarketing
Marketing: Announce feature through our social media accountsMust HaveMarketing
Start PR for this featureMust HavePR
Build Marketing Collateral for partner restaurants in form of tents/ desk with Swiggy branding which will act as kiosks for orders placed on SwiggyMust HaveMarketing

User Flows :

This is how the screens would look

Shoutout to Shivang for helping me out with this.

Mockup 1 Mockup 2

Roll Out Plan:

ItemsStatusDue Date
Testing should be done on design prototypes on both user/merchant end to see what kind of flow is suitable.Done15th Aug
Will start an internal rollout around Swiggy HQ/swiggy employees and restaurants to get clear actionable feedback..Due25th Aug
Post feedback and after iteration, launch to around 50 merchants in 3 populated hotspots, where restaurant to consumer distance is less and volume is highDue1st October
Gradually scale basis marketing events and other promotions keeping in mind #of restaurants > X AND # of consumers who've placed an order > Y in localities, where X & Y can be estimated after pilot run responseDue1st November