Situation
I joined the mobile App team in Q4, 2022. We managed/optimised Android and iOS versions of the app, built in React native.
Problem
The app contained an acquisition journey for a highly profitable product — HomeCare. This had the potential to generate £400 revenue per customer/per year. Over a 6 month period, 98% or ~402,000 customers dropped off at the entry point into the journey.
Goals
Research and optimise the HomeCare journey.
Run experiments to determine if the app is an effective sales/acquisition channel.
My Role
Led the discovery phase through collaborative sessions to generate ideas.
Structured + facilitated a design sprint, which motivated and aligned the team.
Strategised + refined ideas with key team members.
Practiced and promoted data-driven/evidence-based design throughout.
Planned and conducted unmoderated research through UserZoom.
UI and interaction design with Sketch, Zeplin, Marvel, and Framer.
Complication
To satisfy the business request, there were various complexities to negotiate.
Opposing views numerous stakeholders believed the app should not be used as a sales channel.
Configuration the app was not configured or built to enable promotions to be surfaced, without going through the release cycle.
App Store release releasing to Google Play and the App Store is long and labour intensive.
Alignment working with and aligning different teams and specialists from across the business was challenging at the best of times.
Unavailable data unknown truths - do app users complete journeys on the website?
Outcome Expectation
To align different teams from across the business, to investigate and drive product and services sales via the mobile app, to determine whether it is an effective sales channel. If the app was not considered to be an effective sales channel, it's future may have been limited as discussions for a new energy app were already underway.
Remedial Action
HomeCare journey
This conversion/fallout visualisation clearly showed where in the journey visitors were dropping off.
Why were 98% of users dropping-off?
We launched a survey triggered in the app, to learn more about user behaviours and expectations. My ambition was to also recruit for future research — get closer to app users and involve them in our design process. We managed to recruit — over 6,000 active users.
I wrote research goals with the squad in a similar format to a user story. We need to [know this] so that we can [do this]. This ensured we were able to strategically act on the insights we gained.
Survey Findings
Research suggested that product selection in (A) was not efficient or clear. I Simplified this by compressing 3 screens in to 1 (B). We also discovered that product comparison was not easy.
Comparison made simple
A Framer prototype showing a proposed route in to the HomeCare configurator, which offered an easier way to compare the different HomeCare products.
Effective sales/acquisition channel?
Design sprint to the rescue
I was brought into this project to help the team improve on ways of working, and accelerate towards business objectives. A design sprint created ideal conditions to collaborate, rapidly learn and deliver.
Professional speakers discussed:
The business challenge
Data insights
Journey steps
Planned marketing activities
Technical limitations
This provided structure for the wider team to learn, explore, and ideate. Ultimately it aligned anyone with an interest in this project to help decide where to focus our efforts.
Inspiring Demos
Examples of X/Up sell opportunities the group came across in other apps. This fuelled inspiration and possibilities.
Where the magic happened
We used the ‘Crazy 8’s’ method to sketch/form ideas, which were then pitched to the group and voted on.
LEAN + profitable
Ideas plotted on a time/value graph, helped the team to align on what could be achieved in the short term to deliver the most value.
Two days into our design sprint, the team was aligned on our challenge, aware of low performing journeys, and we had generated a bunch of ideas for exploration.
Chosen Idea
Near instant app updates: With Black Friday looming, we decided to focus our efforts on a preferred idea that would enable promotions to be easily surfaced in app.
Hypothesis: We believed that introducing various triggers within the app, that linked through to an in-app marketing screen, promoting a Black Friday offer, would contribute towards sales targets. We’ll know this is true by monitoring the Net conversion figure and total sales.
By reusing existing components, app development was kept to a minimum. A new menu item was added to the navigation drawer, and a quick link was introduced to the home screen carousel — to funnel interest into the marketing screen.
End Results
Reduced drop off rates
We simplified and improved curation of the different HomeCare products, and significantly reduced the drop-off rate by ~35%.
Marketing Capability
We built the capability to show promotions in app, circumventing the release process and configurable via a CMS and marketing platform.
Outcome: Over 13 days, the app contributed to 334 HomeCare 4 sales, which generated ~£134K in revenue. Promotions could now be surfaced in-app without going through the release process. This brought huge cost/effort savings.
Conversion figures over a 13 day period.
What we built paved the way for other promotions to be surfaced in app as these examples show.
Another Experiment
Hypothesis: We believed, when a customer is on the booking confirmation screen, showing a selection of products that engineers can install on the day of their appointment, will provoke interest. We'll know if this is successful by monitoring product selection and the CTR.
Outcome: In Jan 2022, 9,676 unique visitors landed on the booking confirmation screen. 1,497 of these checked one or more boxes, then clicked 'Tell Engineer'.
Summary
Optimised the landing screens for HomeCare products which reduced drop-off rate by ~35%.
Proved the app was an effective sales channel, ~£134K of revenue generated in 2 weeks.
Built the capability to surface promotions in-app via a CMS and marketing platform which reduced OpEx.
Recruited 6000+ active app users as research participants.
Improved how we worked as a team.
