Improving article sharing
Sharing articles is a powerful action for bringing new users into The Post’s ecosystem driven by social interaction rather than marketing spend. This project focuses on increasing the amount of users’ that share articles and increasing subscribers’ gifting of articles.
Role
Lead product designer, led end to end design, research and handoff
Teams
Product, product design, subscriptions and engineering
Timeframe
August to September 2023
Before
After
The Brief
Washington Post readers are our best marketing devices to bring other readers into our ecosystem. Right now the share call to action is only presented to users in the action bar, forcing users to seek that action out rather than nudging them to when they are most apt to share. Subscribers and non-subscribers have different share capabilities - gifting is available for subscribers, and share is available to anyone.
Sharing is a powerful action that feeds a viral loop of bringing users to our own platform to consume content, but we only prompt users to share our stories right at the beginning of a story under the first article image. We can grow viral discovery of our stories by improving signals around sharing at the right moments in the user journey to push more users to share more often.
Early Research / Context / Competitive analysis and explorations
To kick-off the project, I did a brief internal audit of where the share action currently exists on The Post’s site and the current user flow for sharing for both subscribers and non-subscribers. I also collected any metrics for the article page that could be relevant for a new share experience, I found:
Not many users make it all the way to the end of articles, drop off increases as scroll depth increases.
Some users share without reading if the headline is interesting, whether to social media sites or to friends and family.
A few years before this project, The Post’s site had a persistent share button that followed the user as they scrolled the page, but was removed due to disrupting users’ reading and changing visual aesthetics.
The Concepts
I created a selection of concepts that would address the goal of increasing sharing with different strategies organized from lowest to highest amount of engineering lift for product design, product and newsroom leadership to evaluate.
Elevate share call to action the article action hierarchy
Reasoning: share and gift actions already exist in two places on our article pages. Rather than creating more possibly unnecessary opportunities for readers to click to share, we can elevate those buttons place in the hierarchy by visually differentiating.
pros: least amount of lift, doesn’t introduce anything new to our already packed article pages while elevating what we already have.
cons: if the current button locations or misunderstanding what gift/sharing means are causing users to not interact with share, those foes not solve those issues.
Establish the social value of sharing
Reasoning: Commonly found on social media sites, users are often influenced by the number of people that have already interacted with a piece of content, think instagram likes, reddit voting etc. When a post already has a lot of interactions users are quicker to share the post with their own network because of it’s virality. We could tap into this pattern for our share action by labeling articles that have high sharing engagement and including the number of shares of that article.
pros: Creates a visual pattern with the label and share action that could scale easily to other surfaces (homepage, app, section fronts). It leverages a pattern that has shown to be successful on other sites to increase user engagement.
cons: If there are not many shares on an article, could that dissuade readers from reading or sharing that article.
Create persistent call to action for easy access
Reasoning: Giving readers easy, continuous access to sharing will increase shares by allowing readers to access it any times they need it, rather than only at the beginning and end of the page.
pros: we can create bespoke logic for when to prompt readers with share popovers, nudgeing them only when it would be most effective. It also does not have to be tied to a physical point in an article like an inline module would, and doesn’t affect the article content at all.
cons: with our current mweb experience, it is difficult to create persistent or sticky pop overs because of our softwalls, banners and ads.
Eye-catching inline nudge
Reasoning: Prompting readers to share with a combination of visual interest and snappy language in the body of the article will make the prompt more visible, and readers more likely to engage if we prompt at a time where they’ve read enough to know they like the article without being too far past the drop off point for most users.
pros: we can create logic for when to show readers the inline modile, it doesn’t affect or interact with our soft-walls, ads and other in-article modules. Also, the space for copy text will allow us to communicate a value proposition to the reader to make it more enticing to engage with than a stand alone button.
cons: tied to a specific point in the article, so if the reader never reads to that point they won’t see it. We have many other inline call to actions in the article, so we’d need to come up with logic to help prioritize which readers are shown which prompts.
Introduce a scrolled state for the navigation bar
Reasoning: Currently, the global navigation bar is static and does not change at all from when the user loads the page to scrolling. Introducing a scrolling state for our site navigation bar that would include the most important actions for a reader when they are actively scrolling the page (like share, listen and open app) would increase engagement with those actions as they are constantly in reach when a user may need them without interrupt the reading experience with a new module.
pros: Wed don’t currently have a unique scrolling state for our navigation bar, this is a good opportunity to innovate and combine multiple of our high profile initiatives for the quarter in one. Our current navigation bar would still exist, just only when the user is at the top of the article page. This is also possible research opportunity to determine which actions are most important for users to quickly access while they are actively reading.
cons: This has the largest lift of any of the concepts and introduces a whole new navigation bar. This could disrupt some of the other projects to address these initiatives that the team already has in motion. This change would also require the most buy-in from multiple different teams.
During the presentation of these concepts I hypothesized that a one-fits all approach may not be the best solution to address all aspects of this project, suggesting that a staged and multi-solution approach using a combination of the concepts above would most likely give us the best outcome.
Research
After feedback sessions and critiques we decided that using a staged approach would best address the goal of increasing sharing, starting with the lowest lift item as the first release and moving from there: step 1 - elevating existing share hierarchy, step 2 - add social value to share, step 3 - adapt inline share call to action, step 4 - introduce a scrolling navigation state.
I conducted two user tests for this work, first a test for the inline call to action to evaluate what language would be the most effective in enticing users to engage with share in collaboration with our team content designer. The second was a test on the share pop-over for subscribers to evaluate the clarity of the differentiation between gifting an article and sharing an article when both actions exist within the same popover.
During the user test we found that users, contrary to our hypothesis, reacted negatively to the word “free” often confused that they would normally have to pay for the article or that they didn’t have an active subscription. Due to this, we found that option 7 performed best for clarity and moved forward with this design and content.
Final designs
This is where I am going to describe the final decisions and put more in depth images and videos of the final designs
Stage 1
Stage 2
Stage 3
Stage 4
The Results
Stage one and three launched, two and four were deprioritized and currently in the engineering team backlog. The team launched an a/b test with the two new designs of share to evaluate whether to take the new experience to full launch, after the test concluded it was determined that the new designs increased the rate of sharing and was fully launched for the whole site. See data below
Readable Pageviews - this is how many article page views are not blocked by a paywall/hard registration wall and thus gave the user an opportunity to see the share/gift buttons. This is not a kpi but more to make sure traffic is allocated evenly since this is a 50/50 split
Total Share/Gifts - amount of times the share/gift button was clicked. This is a helpful metric but we need to look at the rate because the pageviews aren't always evenly split.
Share+Gift Rate - the main kpi. This = total shares+gifts clicks / total readable article page views. The rate at which people use the button(s)
Shares per user sharing - this is a mostly nice to have metric. It just looks at whether the experience led for users to engage/share more deeply than before. You can say on average, users that use the share button share on average 1.72 articles over the course of the test.