Feature Overview

Deliveroo Plus is the first iteration of a subscription product. It is also the first wide scale experiment we've ever conducted, because of the nature of the feature we couldn't run a simple A/B test. We parameters of the experiment were that we launched the offering in 6 UK cities creating look-a-like groups in other "sister" cities in order to measure it's impact. The initial offering of Plus was a two month free trial and free delivery on every order (without cap).



We believe that offering a subscription service (initially giving free delivery) will increase the order frequency and retention of both our ambassador and active customers. By subscribing users will be more likely to order with us over competitors thus gaining loyalty.


Metrics to impact

  • Order frequency
  • Retention
  • Acquisition



  1. We can't use current location as a true determinator as it's variable.
  2. We don't have a recurring payments system in place.
  3. As user needs a valid card even when signing up to the free trial.
  4. Customer service agents need to know how to handle "Plus" customers


Project goals

  • To convert occasional and active users to more frequent users (ambassadors)
  • Reward our ambassadors with a more cost effective solution as they are already ordering 4 or more times a months (£2.50 per delivery)
  • Drive growth by using the free trial as an acquisition method
  • Test the financial elasticity of different price points including both annual and monthly subscriptions




There were two use cases that I was operating under the assumption of, direct and indirect discovery.

  1. [direct] Discovery during an order by offering the _current _ delivery fee free
  2. [indirect] Discovery from the home screen and available to sign up in your account

The direct discovery was trying to make it feel like you could get something free right now to encourage adoption, and indirect was if someone had told you about Plus or we sent out CRM communications we had somewhere to land customers on irrelevant of an order.

Screenshot 2017-09-25 14.48.25.png



We choose to do a quick branding exercise as this was an experiment in only a few regions after all so we could rebrand the product if it was successful. Working with the graphic team we worked up a few lo-fi directions before agreeing on a simple approach.


Design - annual/monthly


Design - Benefits