Why were so many users cancelling their health food subscriptions?
Our client curates boxes of 7-10 healthy edibles that users can subscribe to on a monthly basis. Subscription sales and customer lifetime value had been steadily increasing — until a rapid downturn for 3 consecutive months.
The products hadn't changed, nor had price, quality, volume or traffic.
We hypothesised that if we could manage users' expectations of the product types, size, volume and billing cycle and instil trust in the checkout process, we could reduce avoidable cancellations by 20% and cart abandonment by 30%.
My role as the sole UX Designer
It was my role to investigate why subscriptions were being cancelled by conducting primary and secondary research.
-
I analysed my research and presented the findings to Alex and Tash.
-
Together we prioritised the designs we wanted to implement and A/B test to solve the problems using the Pareto principle.
-
Alex and Tash implemented the solution with the client while I monitored results.
​
Solutions delivered:
-
Redesigned subscription packages
-
Redesigned customisation & add to cart flow
-
Revamped checkout flow (constrained by Shopify's customisability)
What was the data telling us?
The number of return visitor purchases had indeed been dropping over time, despite increased new traffic.
I half expected to see a decline in traffic to explain the decrease in subscriptions, but when I looked at traffic in Google Analytics and the sales data in Glew, I could see that both traffic and sales across all one-off products had been increasing steadily, but return visitor transactions were down nearly 20% on the prior year.
Empathy mapping
Empathy mapping revealed the product was confusing.
Messaging wasn't clear, subscriptions and billing cycles were cryptic and the checkout added little confirmation or clarity.
I conducted a UX review before speaking to users, to get an understanding of what may have led to this downwards trend in subscriptions.
I wanted to see if I could gain any insight into trends — was this a price, quality or usability problem? What were the key points in the buying experience that could mitigate any concerns that could lead to cancellations?
Analysing Qualitative data
Initial research gave us a long list of obvious things we could improve, but we needed to focus on understanding the subscription problem specifically.
My next instinct was to collect some qualitative feedback to understand the quantified data I was seeing. After surveying 10K customers and receiving 4K responses, I had synthesised results into 14 categories of subscription cancellation reasons.
No one dietary type resulted in more cancellations than others; this graph was closely relative to volume of sales per diet type.
We felt a little overwhelmed with how to decide on which of the 14 problems to solve.
I looked for trends as a start, but landed at a number of dead ends. I grappled with questions like:
-
Do we weight problems with service (ie delivery or poor customer service) the same as product problems (taste, quality, amount)?
-
Were there any more cancellations for a certain dietary type than others? Or subscription length?
That's what led us to use the Pareto principle — to compare problems relative to each other.
The Pareto principle states that, for many events, roughly 80% of the effects come from 20% of the causes.
The key problems identified in pareto analysis
-
Financial situation — most users felt they couldn't justify cost of luxury subscription services any longer.
-
Not using all products each month — these users felt positively about the quality and taste of the products, as well as value for money. There were just too many.
-
Was unaware it was re-occurring — users were surprised that fixed length subscriptions auto-renewed; largely a microcopy problem I found in my UX review.
-
Just testing a single box — the company was losing money on these boxes and considering removing the single box as an option. Only 2% resulted in repeat purchases.
Lo-fi Ideation & prototyping
After defining key user stories, we did some rapid whiteboard sketching and voted on the most impactful solutions I would then synthesise as one concept in Sketch.
We were constrained by budget, a 2 week time frame and the customisability of our client's Shopify website & content. This meant there was a strong focus on UX writing and visual affordances related to guiding the user through the existing experience, over a complete redesign.
After purchasing or renewing a subscription product I can...
-
Clearly see what type and size of products are in a subscription box so that I know what to expect before ordering.​
-
Adjust the volume of products in the box to a value that is financially viable for me.
-
Confirm the customisations of my order once added to cart and again at checkout.
-
Understand billing & renewal frequency so that I won't cancel when I receive the next bill out of surprise.
User stories we set out to solve in these wireframes
Trade-off
Prior to testing we sought to validate the need to adjust number of products per box, as the client was reluctant to entertain such a major operational change.
-
A second survey revealed that while users liked the idea, they admitted that ultimately it would not deter them from cancelling their subscriptions.
-
As an alternative, we added a reason for cancellation step to the cancellation flow.
-
This data later served to inform an evidence-based marketing & pricing solution for this cancellation issue.
User Testing
Testing revealed that we'd inadvertently increased cognitive load, despite users getting through checkout faster.
I tested with 7 users in person, first the existing flow, followed by a clickable Marvel prototype of the new flow in order to determine baselines and any usability issues before the build. I set them each 5 tasks revolving around the user story requirements and asked them to think aloud throughout.
The new design took 27 seconds less on average!
Biggest changes made from testing results
Users were confused by the billing schedule vs pay today in the checkout step.
Users felt that if they were purchasing 3 months' subscription of product, they wanted to see the total and pay that upfront.​
Users wanted to see the value of savings from the discounts, not the percentage.
It significantly impacted speed to decide which subscription length they'd choose.
​
Users were overwhelmed with being shown too much information. The progress bar made it feel laborious, so we opted for progressive disclosure instead.
"I have so many subscriptions already and it wouldn't be worth subscribing to say, 5 products per month"
Ex-subscribing customer
The UI & solution
Mobile-optimised progression of steps to select, customise subscription, add to cart and checkout.
How might we avoid cancellations based on expectations not being met?
-
Streamlined entire flow with progressive disclosure
-
Rearranged flow to select box type before length
-
Added typical size & weight of products in each box
-
Increased quality and size of image with representative content
-
Confirm all customisation details in footer
How might we ensure users understand billing cycles and auto-renewal policies prior to purchase?
-
Reinforce billing cycle in footer much earlier than checkout
-
Optimised layout to reduce noise and pull focus on key details that can lead to cancellations
-
Added logos to increase feelings of security and trust at checkout
Results
22% drop in cancellation rate 2 months post live
This result is exciting because the cancellation rate is actually better than it had ever been, even prior to the downward trend!
53% reduction in cancellations related to testing products out
The company used to lose money on one-time customers who cancelled for this reason.
19% reduction in abandoned carts
Heat-maps revealed that users were spending less time on this page, had less enquiries related to billing.
"This is way easier to use. The cart used to look a little bit dodgy, like I shouldn't really trust it"
GMB Health food box subscriber
"I didn't even realise you guys had beauty boxes. i'd love to give them a try too! Love the new subscription page"
Long-time GMB health food box subscriber
TITLE OF THE CALLOUT BLOCK
Learnings
I learned how to sprint — perfect is the enemy of deliverance.
-
I learned how to do things in a scrappy but methodical way so that we could get feedback and ship to deliver the most value quickly, thanks in large part to Pareto analysis.
-
Then once the MVP solution was out, we could iterate and optimise based on user feedback.
-
Update years later: I learned how much UI trends have evolved as well as my own taste and skill, and i'd design it completely differently from this daggy UI these days!