@CHEWY
Frequency recs
Shaping a project beyond the brief
Chewy’s Autoship product is highly successful, however core usabilities issues create a high rate of churn and customer contacts. One of the most difficult parts of the experience for customers to manage is understanding how often they need their order to be delivered.
16.5% of Autoship cancellations are due to “difficulty scheduling”
The problem
Support customers in selecting the right cadence for them
The project
Competitor subscription services recommending a delivery schedule to customers in both acquisition and management experiences.
Project goalIncrease conversion in checkout & decrease Autoship subscription cancellations
Obvious solution
The obvious place to put a recommendation was in the existing placements at Checkout and Manage Autoship. However those placements had some disadvantages.
Chapter 1
Existing frequency placements
The obvious place to recommend a frequency was in the existing placements at Checkout and Manage Autoship. However, those placements had some disadvantages.
Obvious placement candidates for adding a recommendation and matches what competitors are doing.
An observation
Only 1.2% of the 800k+ weekly sessions interact with the frequency picker
Less obvious solution
Invite customers to interact with frequency at opportune moments.
Chapter 2
“Frequency nudges”
Since customers didn’t like, and weren’t used to interacting with frequency, I proposed we intercept them during moments where they were signaling an issue with their current frequency. My proof of concept example was immediately after a customer has “skipped” their order, and had done so multiple times recently.
Early wireframes showing potential intercepts for customers, asking them to engage with their order frequency.
Concern
The nudges will distract customers from the success message. To get signal on whether this concern was valid, I conducted a quick usability test with a prototype. The results were positive, and gave me enough confidence in my idea to begin campaigning to PM that this solution was the right one.
Phasing
PM wanted to launch the most basic portion of this project as fast as possible, believing we were leaving conversion money on the table. Which does make sense!
However, I felt strongly most customers would not even see the recommendation on Manage Autoship and retention would remain flat. I continued moving the project forward, knowing we’d revisit the issue in later phases. My nudges and other ideas (moving frequency selection up the funnel) were pushed all the way to phase 4, the last phase.
Chapter 3
Phase 1 launch
We launched it in the obvious places, Checkout and Manage Autoship. I did however introduce the frequency picker in a sheet component with the long term plan of making it reusable in the upper funnel.
Chapter 4
Phase 1 Hi-FI designs which launched late December 2024
Results
While there was a small increase in conversion (+26bps), cancellation metrics remained flat
Making Phase 4 Phase 2
After our flat results, PM decided that we should move the nudges (what we were calling phase 4 at the time) up in prioritization.
This project is current and ongoing. You can see my next high-fidelity design iteration below.