Tu Help expereince

Reimagining the Help experience for TU


How I expanded a projects scope to uncover a larger problem and create a holistic vision for self serve Help.


There is a large scale migration piece happenign at Sainsburys Argos in a cost saving move (all Products on a single tech stack). As a result of the replatforming there is a need to look at each part of the experience and look for opportunituies to update and improve.

This project was started off with the ask to sense check the UX of the accodion component in the new Design System and how it compared to the current implementation, it ended with a complete reimagine of the Help experience across all brands.

“Hey Chris, can you do the UX for this accordion”
lead to the realisation that the Help section of the site was not being properly considered in replatforming work.


Help is a massive area, there is no way a single team would be capable of creating the right experience for customers for such a complex problem so this required reaching out to both content team and the CMC (Call centre Team) to get an understanding on whats been done so far, whats worked, what hasnt and what is the current direction for Help from a content perspective.

Tu’s current help pages are very information heavy as a result Customers are having to phone up the CMC on average approx 1 in every 5 orders to get help. High number of CMC calls add additional cost to the buiness as well as being a more lengthy way for customers to complete their tasks. There is an opputuniy to reimagine the experiece to allow customers to self serve reducing CMC costs and increasing customer satisfaction due to a quicker and easier experience.

Customers are having to phone up the CMC on average approx 1 in every 5 orders to get help


Discovery continued

Inbetween the X team collaboration work I set about benchmarking our competitors, looking at time on task, confidence and NPS of comparible journies based on Heatmaps and CMC data (what are our customers trying to do the most).

How this was benchmarked (specific questions and amount of people)

Hypotheses: We beleive splitting content into tasks and info will decrease time on task and increase nps


Design goals

Reduce customers reliance on Call centre (drive down calls per order) (Business)

Help customers self serve - completing tasks quicker and easier (Customer)

Increase UX Benchmarking to match or exceed top scoring competitors (Both)


Based on what the CMC and Heatmaps were telling us (and with an eye on high benchmarking competitor references) it appeared the majority of customemrs wanted to be able to complete/find info on three main tasks, pulling these out and creating self service journies made sense so a new Help home page was wired up based on the Hypothesis that seperating the content into Primary tasks and secondary information would benefit the majority of customers as well as iteration on the current flows (becasue we need to define the path into the flows before iterating on current flows).

Wires and FLows


Going back to the replatforming objective the visuals were created using as close to possible the current CMS components available for builds. I iterated slightly on a couple of visual styles to make them more suitable for specific usage (for example reducing the size of imagery to use an icon) in the hopes otehrs would agree on ROI once we got to build.



Benchmarking was done by identifying which tasks are supported on all brands and which are the most used witihin Tu help. Returns and tracking were identified as appropriate tasks as they were x brand.

Users were asked to find out how to return an imaginary purchase across 8 brands (list brands) with the order randomised to reduce any bias through learning. Time on task was used as a measure of success as well as getting users to give an NPS and confidence on task rating.


The test subjects scored the slightly better than the live site and their personal preferences were the prototypes if any. It is rare testing is concrete or predictable with the surprise part of this test being 50% of the subjects not having a clear favourite, or even noticing the differences. This however isn’t necessarily an issue as the goal of the test was to prove the new visual designs produced an at least an equal usability score to the live navigation while enabling the new tech stack.



All content © 2024 | It’s been wonderful