Dora, a social-first article experience for Bloomberg.com

(Launched 2019)

 

How might we better engage social-first mobile news consumers?

When Bloomberg launched its digital subscription business last year, BHIVE honed in on a data signal that’s a chronic pain point to publishers, especially ones that want people to read and pay for premium journalism: bounce rates among social news seekers.

Like many news sites, a growing base of our visitors each month come from “side door” entry points like social and search. While these visitors bring traffic to our site, our deep-dive into the analytics showed us that some of these users are not likely to return within a given month. BHIVE set out to understand why Bloomberg.com’s social article readers were the likeliest of our online audience to “read-and-leave”. To see how we might break away from this “one-hit wonder” phenomenon, we framed our exploration, which we called Project Dora, with a simple question in mind: What do social news audiences want from a mobile article experience?

The findings from this study led BHIVE to prototype and test a mobile web article template designed specifically for people coming to Bloomberg.com via social. In a trial run to 5% of our audience, Dora tripled article engagement (via articles shares) and led to significant increases in article completion rates and decreases in exit rates.


A screen cap of the Dora mobile article experience, a simplified format designed to better meet the needs of social users who engage with Bloomberg.com content as “in-app” experiences via Twitter, Facebook or other social feeds.

“I skim Facebook quickly. If the picture and headline summarize enough, I won’t even click on the article…if I see an article that I’m interested in, then I’ll click on it and like skim over it.

—Canwen, 20, a social-first news seeker who spoke to how she generally catches up on news


My Role

Researcher, team lead


20% of Bloomberg.com users were coming to the site via social on mobile devices monthly. These users exited at higher rates and were likelier to be “one and dones” than all other users.

Our hypothesis: the existing Bloomberg article experience isn’t designed to take into account how social users engage with news content.

The Challenge


Research Solution

  • User research (competitive testing) — understand how users discover and engage with content via social

  • Collaborative ideation and prototyping — identify ways to improve the article UX for social

  • Usability — test and provide recs for refining UX/UI prior to MVP experiment launch

  • MVP experiment — identify an MVP that could help to validate the concept and inform future iterations

 

Findings

Discovering our audience

First, we analyzed our mobile web traffic. At face value, high bounce rates are considered a sign that a website isn’t providing necessary content users want, but our research offered slightly different key insights for understanding the bounce rate problem as it pertains to social news:

  • One-fifth of Bloomberg.com’s monthly visitors came to the mobile site through article pages shared via social sources

  • Bounce rates for this audience were higher compared to non-social visits

  • Social visitors were likelier to be “one and dones” who came to the site once and did not return within a month

The numbers provided insight, but to understand the Why behind the What, we sought to learn directly from social news seekers. Through a series of qualitative studies, in-person and remote video interviews from across the country, we sharpened our perspective of their news expectations, including their unmet needs. Some of our takeaways from the research:

[Experience]

Social reading is a highway for skimming and scanning — articles are the detour.

Our takeaway: Skimming a home feed is an exercise in quickly taking in the news and information scenery. Clicking through to a link requires ditching the social fastlane. It should come as no surprise that people are discerning about what they’re willing to “pull over” for — news articles are no exception. The slightest friction — pop-up ads, slow load times, desktop UI crammed onto a phone screen — are signals that they should merge back into the social fast lane.

What we observed: We observed social news seekers who wanted to stay focused on content that stood on its own within their feeds — social cards that included animated GIFs, short videos with captions, pithy or interesting posts with an immediate insight or takeaway. When people did click through to an article, their natural tendency was to boomerang — or read and return — to their social feeds, not explore the host site for more news.

[Expectations]

Clicking through is an act of validation, not exploration.

Our takeaway: By the time social readers reach an article, they have context. Whether it’s expert commentary via tweet or a joke riffing on the news on Reddit or Facebook, a good social post has usually stolen the punchline of a news article before the reader leaves his or her feed. As a result, social articles are less about reading a story from beginning to end as they are an exercise in corroborating a takeaway or opinion formed from within their social feed.

What we heard: As Canwen, 20, a NYC college student put it: “I skim Facebook quickly. If the picture and headline summarize enough, I won’t even click on the article. If I have more time and if I see an article that I’m interested in, then I’ll click on it and like skim over it. I won’t read over in-depth, but I’ll skim over it for a minute or two.”

[Engagement]

Social sharing doesn’t mean putting your opinions on full blast.

Our takeawayNo one wants to be that person, the one who puts way to much out there on social media. Our conversations with social news seekers pointed to much more subtle sharing habits: Private group or one-to-one conversations are the likelier route for sharing among the social news seekers we interviewed.

What we observed: Re-posting and liking articles is public to your followers and social network, (e.g. RTs, likes, comments), and therefore many social news seekers do so with a discerning eye. Instead, we saw people going out of their way to copy, paste and share articles to friends directly via text and chat platforms like WhatsApp and Slack.


Product Deliverable

A before of the Bloomberg social article experience.

A before of the Bloomberg social article experience.

The Dora redesign introduces an overall reduction of key elements: smaller typography in the headline, less prominent branding, removal of story bullets, and shorter bylines and timestamps. Plus, new features: a fun and irreverent tone brought to li…

The Dora redesign introduces an overall reduction of key elements: smaller typography in the headline, less prominent branding, removal of story bullets, and shorter bylines and timestamps. Plus, new features: a fun and irreverent tone brought to life through tiled (space-saving) photos and a floating, thumb-friendly copy/link button. (Design by Tiff Hockin)

The final result was Dora, a social article experience that directly responds to user needs, as uncovered through our research and prototyping, including:

Mobile-first experience: optimizing our design and experience for small screens made our content more accessible to social news seekers on the go.

Social-first article template: our simplified article template cuts out the clutter typical of mobile websites, getting social news seekers the content they came for faster and reducing the emphasis on stuff they don’t care about.

Native social newsstand: Dora introduced Bloomberg social readers to Open Graph protocol, an upgrade to our tagging system that made our article links on Twitter and Facebook more visual and easier to click through a link.

Floating copy link: a bottom-right copy button makes it easy to grab and share links with minimal effort.


  • Users shared 3x more — clickthrough rates on the “copy link” button were 219% higher than all other social share buttons combined

  • Users scrolled deeper:

    • Pageviews where 25% of the article was viewed increased 4.1%

    • Pageviews where 100% of the article was viewed increased 7.4%

  • Users exited the page less often — exit rate from article pages decreased 5.8%

  • More ad views display ad impressions per user increased 8.7%

  • Successful MVP refined and launched to 100% of mobile app users in 2019

Impact