Since I conducted the UX Audit back in 2016, I have been retained by Mulberry as their sole UX Design Consultant. Over the past two years we have established a UX improvements programme.
The main purpose of the programme is to increase customer satisfaction, brand loyalty and conversion. One key focus is the mobile user experience.
This case study will showcase how the mobile navigation has been redesigned with the aim of improving usability, discoverability and conversion.
Google Analytics can be a veritable goldmine of information. Event tracking, goals and heat maps can be especially helpful when analysing user behaviour. Unfortunately many of these features hadn’t been set up for the mobile navigation. So we were forced to resort to slightly higher-level metrics.
Above is a breakdown of traffic for the month of May 2016. You can see that more than half of all traffic is from mobile devices, although with just a 0.14% conversion rate, this only accounts for 30% of e-commerce revenue.
The 2016 UX Audit highlighted an overall score of 68%. This indicated that the existing mobile navigation and search suffered from several problems:
Creating a hypothesis enables you to quantitatively evaluate the success of your design. This enables you to reinforce design decisions with data rather than just preference or expert option.
Using the insights from both the analytics and UX audit we defined our hypothesis as:
By improving mobile navigation, users will be able to find product information more easily. This will result in lowering the bounce rate and increasing the percentage of successful transactions on mobile devices.
Below are three screen grabs showing the pre-existing Mulberry mobile website.
It had a confusing navigation structure, no affordances to indicate to users that there were sub-navigation options available, and a back button positioned on the wrong side (right hand side).
Navigation level 1 showing main categories for Women, Men, Collections and My Account.
Navigation level 2 showing the sub-categories once the user has selected 'Women'.
Navigation level 3 showing tertiary level options within 'Women's Bags'.
To test our hypothesis I undertook a number of specific design changes. These included:
Adding arrows next to navigation items containing sub-categories. This ensured that users could distinguish between these and normal links and thus improve discovery of the sub-category items.
Additionally I increased the font size for all links to make the navigation easier to scan and help users locate the option(s) they were looking for more easily.
Low-fidelity sketch of the new mobile website navigation.
Hi-fidelity wireframe created in Sketch showing level 1 of the new mobile website navigation.
Animation showing the spatial design for the existing mobile website navigation.
Well-designed mobile apps use considered movement and spacing between elements to create a spatial model that helps the user orient themselves within the experience.
The existing navigation had a poor spatial design for the different menus within the hierarchy. All elements were stacked on top of each other and just being shown/hidden when the user interacted.
This coupled with a lack of animation, resulted in an abrupt feeling when the user toggled between states.
Understanding what's currently in the viewport, what has left the viewport and what can be expected to come into viewport when a user interacts helps the them orient themselves within the experience.
I redesigned the horizontal placement of the menus within the hierarchy and combined this with simple transitions between states. This created an improved spatial model for users and improved usability.
Animation showing the new spatial design for the mobile website navigation.
I used Flinto For Mac to create a high-fidelity prototype using the static assets that I created in Sketch. The prototype was used to communicate the behaviour of the UI to internal stakeholders within Mulberry. It also acted as a visual reference point for the external development team to develop their codebase.
A screen shot of the Flinto For Mac interface showing the main artboards and the connections between them.
It’s been two years since the mobile navigation has been updated. Comparing the same top-level metrics from May 2018 against May 2018 indicates that, on paper, there redesign was a success.
Although the data above points to an improvement, there could have been other factors that have influenced the outcome. Unfortunately due to the absence of specific event tracking and goals, we couldn't isolate the mobile navigation metrics as well as I would have liked. This is a key learning for future projects.
In addition to improving our event tracking and goals in GA, there are some additional tools that I have incorporated into our design process:
We use the heatmap and recordings functionality to measure levels of user engagement.
We have started testing certain module variations on PDPs to determine which creative execution delivers the highest level of engagement.
The HEART framework helps provide me with a consistent approach to defining KPIs as part of the early phases of UX strategy.