Having grown the unique and repeat editorial page views as well as viewability and impressions of the adverts the business relied on we knew that replacing our out-dated, RSS-fed native apps could increase loyalty with our users, meeting their expectations of major national news titles and improve targeting for our commercial team. The challenge was to build current, engaging apps to compete against some of our huge competitors in the industry.

Our proprietor and directors were directly invested in this project so the pressure was on and myself and the product manager engaged the digital editors and head of commercial and programmatic advertising at an early stage to ensure our requirements captured every opportunity for the business as well as our readers. We employed an external development team for delivery and their input from initial possibilities to finishing touches was invaluable.

Our user base was huge and budget (and appetite) for UX research non-existent so I initially looked at the behaviour of our mobile web users to learn more about their journey patterns, dwell times and routines and identified  affinities to create broad archetypes to represent particular target groups we could devise a strategy to engage. 
‍
From the data we compiled using GA and Hotjar we aggregated characteristics to create archetypes for Express and Daily Star readers. I'm illustrating Express archetypes, though in reality we found that online behaviour, very differently from print audiences, was much more title-agnostic than we had initially thought particularly on certain topics like sport, gaming and tech.

• Express viewpoint news - These were users who regularly visited the site at predictable times of day and week entering through organic views of the homepage so considered brand-loyal. They viewed higher than average pages per visit, particularly focusing on a number of news topics including EU focused UK politics and Royal news. They also regularly viewed image galleries.

• Live events - These users responding to SEO performant live stories, particularly around live political, celebrity, terrorist and football news. They were strongly biassed to mobile devices and responded to aggregated breaking news formats where live-articles were actively internally linked to established entities.

• Celebrity and entertainment news - These users engaged with popular SEO-performant stories, inbound from search engines, aggregators and social media. They were primarily mobile users and consumed more image and video views than other segments. They also bounced more quickly and showed less title-loyal behaviour. 

• Gaming news - These users typically arrived following live gaming news stories (network downtime, in-game offers etc) then showed increased engagement with tech and gaming news and commercial content. This segment was particularly commercially valuable.

• Premiership football consumers - A huge audience with much competition where we had most success with live news stories over match coverage. Users showed some brand-loyal behaviour but more importantly were team-loyal, consuming relevant content around the players, performance and stats.

The initial design stage was to agree the content architecture with product and editorial to align app feasibility with the CMS that editors used to control article and section flow on the website. This was crucial so the curated experience in the app had parity with the web experience capitalising on the skills they had engaging their readers, especially when live events were unfolding.
‍
One of the criteria for the project was that the core product could be re-skinned and adapted for both Express and Daily Star so all our decisions included retaining required flexibilities with common functionality. 
‍
Another essential factor was balancing loading speed with immediately available content. When a user launched the app we wanted to ensure it cached a satisfying amount of appropriate content so if, for example you were jumping on a train you would be able to read headline stories without having to download them separately. This subtle behaviour significantly impact reader experience so we tested and selected the number of news, football and entertainment stories that we thought best balanced these requirements.
‍
From a number of alternative concepts we agreed a structure closely aligned with the web experience but with some key differences to reflect behaviour of previously defined archetypes. These included surfacing popular sections more prominently in the flow of the home page to showcase curated content in key areas, changing the navigation options to promote common journeys for these readers and adapting flows within specific sections such as football and entertainment to reflect popular topics. We also ensured changes in the editorial article and section structure on the site were propagated into the app automatically from the CMS to maximise valuable editorial control. 

Having developed the visual languages of the online brands, aligning the experience across the web and app experiences was a priority for me so I sought to adapt styles rather than reinvent them. There were still many decisions around type styles. Image, gallery and video behaviour that needed to be addressed, some of which had to be designed to replace inherited browser behaviours. 

I designed and tested Invision prototypes of these key elements with teams across the company and with our lack of testing resource even ended up posting images of text styles in the corridors for people to tick-vote which helped me decide between similar, equally fitting options for his crucial detail which I felt deeply communicated the desired authority and accessibility of the content. 

Following the very detailed and lengthy QA process across all delivered UI and transitions (including personally creating custom alert sounds) I designed and delivered a suite of rich animated ads and content assets to drive downloads for our relaunched apps for both titles.

We launched the apps just before Christmas to hit the download spike seen over previous years (with gifted devices and time at home) and immediately hit those acquisition goals. Over the next few months we reviewed feedback and behaviour and made some changes to recommended article devices to optimise second page views. We took many learnings from the launch but ongoing optimisation became difficult to prioritise. Key learnings were that app behaviour varied considerably from online behaviour and that selling commercial opportunities against this audience required a topic segmented strategy that we were just beginning to invest in.