Modernising MySpark Business self-service on a legacy platform

Picture an office manager who needs to add a new staff member to their company’s Spark account. In the old portal, they couldn’t do it themselves. They had to fill in a form or call Spark.

MySpark Business, earlier called MySpark Digital, was Spark’s portal for business and government customers. It covered mobile, data, broadband and managed services, with thousands of forms and variations across products and customer types.

Most tasks ran on service requests. A customer filled in a form, or called the contact centre, and Spark staff did the task behind the scenes. Key features sat three or more clicks deep, and some customer groups had no digital option at all. For Spark, that meant support costs that grew with every new customer.

I was the only UX designer on MySpark Business, and I set the design direction for the portal for over two years. I joined Spark as an Interaction Designer, and my title became Senior Experience Designer after a restructure.

My work covered the whole portal: navigation, existing journeys, moving service requests to self-service, new products, specialised micro-portals, buying and managing services, and login, users, billing and usage.

• I owned discovery, journey mapping, user flows, interaction and visual design, prototypes, customer testing and handover.
• I worked as a trio with the Product Manager and a Senior BA, supporting up to four squads at once, and planned journeys four to five sprints ahead.
• I worked with CX, engineering, sales, account managers, billing, contact centre staff, network engineers and provisioning teams.
• I mentored three designers who joined in later phases.

It would have been easy to start with what looked broken. Instead, we focused on why customers were calling Spark, and which tasks could realistically move to self-service.

The research covered both sides. More than 30 interviews with small business, enterprise and government customers. Usability testing in Spark’s UX lab and remotely with regional customers. Interviews with the call centre, sales, network engineering and provisioning. The analytics and research agreed: customers couldn’t finish basic tasks without help.

Making the problem visible

The problem crossed a lot of teams, and most had never worked with a UX designer. So I built a wall. Using Spark’s JUCCI framework as a guide, I laid out customer needs, the core journey steps, pain points and questions, detailed user flows and the existing screens.

People from sales, billing, engineering and provisioning would stop, ask questions and add what they knew. Those conversations connected customer needs with the processes and systems behind them, and approvals got faster because people already understood the problem.

Examples included adding a mobile plan, changing a plan, and adding data to a corporate connection. For some of these, about half of customers chose self-service straight away, while the rest kept using service requests.

Prototypes were built on Spark’s design system and went through more than five rounds of changes. We only took a journey forward once it scored above our 80+ testing threshold. I handed over annotated specs, edge cases and WCAG AA notes, and stayed involved through design QA.

I designed a connected purchase journey for devices, plans and accessories. Customers could start with any of them, set up what they needed, and add related products before checking out.

Design was one part of these results, alongside engineering, product and support.

The wall became the standard way to do journey mapping across Spark business portal work. I added business patterns to the design system, and received the Best Team Player award for bringing other teams into the design process.

The wall turned out to be more useful than I expected. It gave teams a shared object to work around, and most of the alignment happened in front of it, not in meetings. I’ve used big, visible maps the same way since, whenever a problem is too big to hold in a conversation.

I also learnt not to wait for perfect technology. Working within the old platform got real improvements to customers far sooner than a rebuild would have. That office manager could now add a user without picking up the phone.