Project Details
Duration – 1 month
Team – 1 software engineer, 1 product owner and me
My role – Created MVP and iteration designs, ran user testing
Tools – Sketch, InVision, Photoshop
Deliverable – MVP and experiments designs
Project Objective
To maximise revenue from our existing dry food range by offering greater value and choice to more dogs and their owners.
Hypothesis
We believed that:
Assumption A: Some users will pay a premium for dry food that offers only the highest quality ingredients and a wider array of health benefits.
Assumption B: Showing clear ingredients claims for each blend will grow a user's confidence and increase their perception of value.
Assumption C: The value of a standard tier will increase when displayed alongside a premium tier.
Lean UX Approach
We took a rapid validation approach by running a series of A/B tests to validate the most successful idea before rolling it out to 100% of users.
Measures of Success
Dry ARPO growths
Sign-up metrics remain flat or grows
Claims Survey
Before starting the design process, a survey was sent to our existing users to better understand what claims are the most valued when choosing dry food for their dog. The claims results were mapped to different dog segments (healthy dogs, dogs with health issues, large breeds, etc) to offer more tailored options.
'High in meat content', 'flavour' and 'healthy skin and coat' were the most voted claims by healthy dog owners who were the primary tails users.
Ideation Workshop
The product team brainstormed on how to meet more dogs' needs by offering dry food choices, taking into consideration survey results and other shared insights.
MVP Design
I took the most voted ideas and turned them into wireframes and high-fidelity designs. Two options were chosen for an MVP — 'base' and 'skin and coat'. We ran these designs through some basic qualitative user testing to capture any pain points before rolling them out as A/B tests.
Wireframes
MVP high fidelity designs
Qualitative User Testing Results
The first options seemed ‘basic’ ‘generic’ ‘for most dogs’ and the other two options seemed ‘special’ (not always in a good way as special can mean ‘not for me and my dog’).
The two options didn’t feel tailored - 1/4 thought all users see the same recipes and 3/5 assumed each user sees something different, with users often reporting that the tailored options were not relevant to them.
MVP Design Iteration
Based on the feedback, I iterated on the MVP designs to make them feel more tailored by:
Presenting a playback of information provided about the dog
Visually aligning both options to look equally important
Experiment 1: MVP 'base' and 'skin & coat' blends
This design was rolled out to 25% of the user base, leading to an increase in ARPO while sign-ups stayed flat.
Experiment 2: 'base' + 'gluten-free' blends
We launched another experiment with 'base' and 'gluten-free' options to 10% of our users. There was a 3% drop in sign-ups therefore the test was paused.
I investigated why some users were dropping off from the sign-up flow. The main hypothesis was that healthy dog owners did not feel that gluten-free recipes were tailored to their dogs' needs which led to a drop in confidence and them leaving the flow altogether. This also validated survey results where healthy dog owners did not favour 'hypoallergenic' and 'gluten-free' claims.
I also investigated different metrics such as 'scroll reach' to better understand how users were interacting with the designs to optimise them.
Scroll Reach Heatmap
