EE
I worked with EE on an exciting new project. Creating a tailored checkout experience for existing EE customers looking to upgrade or buy additional products/services.
Previously existing customers were pushed down the same journey as non-customers and forced to enter all their details again. By automating key parts of the journey we were able to create a personalised and expedited experience.
The challenge
The driving force behind the project was some really alarming data:
- X amount of existing customer sales
- Length of flow
- Number of automation orders
Serious issues with the backend systems and processes meant phone agents needed to step in to complete the vast majority of orders, which was extremely costly to the business.
Customers had to re-enter information about themselves that EE already knew, making the journey unnecessarily long. Worse still, the poor online credit check service meant that even after completing the checkout customers were being called up and told they simply couldn’t afford the phone they believed they had already purchased.
THE GOal
The goal was to get existing customers to identify themselves so they would only see the plans and products they were eligible for. We could then make recommendations based on their usage history and product preferences. Being authenticated also meant that we could automate parts of the checkout and expedite the journey. The key customer and business goals were to:
Significantly reduce the number of orders needing manual fulfilment.
Improve checkout conversion.
Meet cross sell targets.
Increase logins and registrations to My EE.
the process
The project was separated into three work streams; Login/authentication; Personalised Shop; and Checkout. I lead the experience for the largest stream; Checkout.
The end-to-end journey would take customers through different areas of the site such as Shop, Checkout and My EE. Similar to most large organisations, EE’s online sitemap mirrored its internal business structure. Meaning each site area had dedicated teams of developers, designers and product owners resulting in multiple stakeholders for me.
In order for me to create a consistent and complete checkout journey it was crucial for me to understand the wider end-to-end journey. Therefore the first thing I did was to map it out, talking to stakeholders to understand the entry points and overlap.
I then facilitated a card sorting exercise with key stakeholders to help learn the desired content order. I then overlaid this with funnel data from an existing upgrade journey to identify the potential pitfalls.
I was then able to create user stories for the main journeys, highlighting unhappy scenarios and key process points.
Moving on to wireframing. I stitched screens together in Marvel to create a simple prototype and then shared it with the project team to get early feedback.
The low fidelity nature of my designs meant they could be quickly adjusted. Iterating through stakeholder reviews and guerilla testing.
The business were keen to increase product uptake and looking for ways to introduce cross-sell within the upgrade checkout flow.
Cross-sell took two forms; first add-ons, extras that could be added to the plan like international calls and insurance. Second being the purchase of an additional products/ plans usually for a family member.
Non surprisingly the latter presented the biggest challenge. I was reluctant to include this type of cross-sell within the checkout flow with fears it could impact conversion. My instinct was that it was better to allow the customer to complete the upgrade before distracting them with what could be a complex proposition.
I created two upgrade journeys: cross sell after confirmation (option 1) and cross sell interstitial (option 2) then tested both with real customers. I worked alongside another designer to create a high fidelity and fully interactive prototype suitable for user testing.
user research
I worked with an internal researcher to identify the journeys we wanted to test and she liaised with SimpleUsability to conduct the usability testing. Sixteen users, with an equal split on desktop and mobile, were recruited to take part in moderated sessions that would include eye tracking and a post task retrospective interview.
The main objective was to observe users on the EE website prototype containing upgrade options 1 and 2. They were counterbalanced across subjects so that users only saw one upgrade journey. They were given micro-tasks to test keys sections of the interface as well as their overall understanding of the experience.
Most users who saw the interstitial version (option 2) of the cross-sell thought that it was in the right place as it could be done in the same transaction, and was not too disrupting if they did not want to add it. Six users who saw the cross-sell on the confirmation page (option 1) thought that it appeared too late in the journey as they would not want to think about another purchase at this time. Therefore the interstitial was the option we chosen to build.
