Sky Go: Homepage customisation

Background:

Sky Go is the companion mobile app for Sky TV products: Sky Glass and Sky Q. Customers can watch on-demand content and live TV, alongside setting and watching recordings.

Product research with had shown that customers felt that the app wasn’t personal enough to them, and there were opportunities to improve this.

My Role:

I worked as the only UX designer on the project, and collaborated with product leads, other UX designers, developers, a UX writer and a UI designer.

Work included competitor analysis, stakeholder interviews, data analysis, ideation, prototyping, unmoderated research, presentations to senior stakeholders, design, and dev handover.

The Sky Go App

💡 Discovery

Some early discovery activity included:

🛒 Competitor research
📈 Data gathering
📱 Audit of current app and homepage structure
🧑🏻‍💻 Desk research into customisation, personalisation and best practices
🤔 Gathering questions and unknowns

📝 Workshop

I organised a workshop during our discovery stage with stakeholders to:

  • Better understand the technical constraints and project scope

  • Challenge some aspects of the brief and better understand the business goals

  • Put forward some initial desk research and align on some principles

🧠 What did we find out?

🖼 Context

  • There are no ‘user profiles’ on the app. Each device is treated like a profile, so customisation changes would only affect one device at a time.

  • One account is usually used on multiple devices, by multiple household members.

  • Part of the goal of the project is making the app ‘cold start’ more personal, so new customers arrive at a unique homepage.

  • Research was showing customers linked poor ‘personalisation’ to actively seeing particular content categories like ‘sports’ and ‘kids’ on the homepage, if they didn’t watch or subscribe to them.

🤖 Constraints

  • Machine-learning and algorithmic homepage personalisation was being explored with other teams, but was not part of the scope of this project.

  • Effects on the homepage would be limited: removal of rails, prioritisation of content. Rails and content are currently manually merchandised and scheduled.

✍️ Sketching, thinking and ideation

 

🪜 Definition of stages

As we iterated and experimented with ideas for different screens, layout and copy, a larger ‘narrative structure’ began to emerge. This enabled us to give a clear rationale for each screen, consider the users needs at each stage and begin to define the navigation logic, along with some draft copy.

 

🧪 Testing

After we had designed and iterated on the onboarding journey, and taken the approach to senior stakeholders, we wanted to bring the journey into testing to validate and observe:

  • The overall usability of the journey

  • Users perception of time taken

  • Users expectations of effects on the homepage, and comprehension of the tasks

  • An ‘all upfront’ journey vs a ‘some later’ journey (as a randomised A/B test)

  • How existing customers fared vs new / non customers.

I set up unmoderated tests, and recruited two groups of 100 participants, on our research platform Userzoom. The test prototypes were built in Figma.

👉 Testing Results

I analysed the test results with some support from the UX researcher, pulling the quant and qual data in miro, so we could see the broader picture, pull out insights and it could be easily accessible to product stakeholders. Some findings were:

  • There were no major usability concerns with the journeys. Both sets of participants completed them quickly, and, when asked, showed they correctly understood what each stage of the journey was for.

  • Significantly more participants preferred the ‘upfront’ journey.

  • Attitudes of ''going into granular detail” 93% of Sky customers and 73% of non-sky customers chose ‘I think it’s worth the effort’.

  • Feedback on comprehension of the ‘granular’ category buttons, which were in popularity order. These were subsequently updated to alphabetical.

 Final Journeys

  • After agreeing on a direction with product owners and stakeholders, I finalised the UX and created handover specs for all versions of the journey, including the logic of skipping stages, and re-customisation – incorporating the feedback and insights from our testing stage.