Something weird happened in 2017. National publications, in the midst of a booming readership year, abruptly began bleeding pageviews. On Facebook, story shares dropped steadily for months, then plateaued — and stayed there.
The cause was something small and technical: Facebook had tweaked its “feed” algorithm, from prioritizing news publishers in users’ feeds to prioritizing individuals. But the implication was astonishing: A small algorithm tweak by a social media platform could radically shape the readership of a publication, by influencing who saw which stories, and when … or whether they saw them at all.
Far from the days of watching nightly news or getting a newspaper delivered to our door, we now rely heavily on third-party curators to serve us the news. Third-party curation has been on the rise since the 1990s, when news and opinion aggregators like Yahoo, HuffPo, and The Drudge Report began culling and opining on stories from a variety of sources, journalistic to sensational. But when Facebook and Twitter introduced timeline feeds in 2006, social media developed the capacity to aggregate and disseminate news across its own platform at scale — with the added benefit of a substantial built-in user base of individual readers, recommenders, and sharers.
Today, what used to be a system of story dissemination directly from outlet to user has become a sprawling system of peer-to-peer shares and constantly-updating aggregators. Mobile aggregators like Chrome’s Articles for You are fast rising, but Facebook has historically dwarfed the rest: As of this year, Facebook alone still counted for 65% of referral traffic.
The implications of this for publications are clear: From a business perspective, it’s risky to lose control over how your product gets to your customer. It’s even riskier when much of this control currently rests with one third-party company, one that has in the past used inflated reporting in an attempt to get publications to restructure to video, and until recently refused to pay publications for the content it was distributing.
From an editorial perspective, it’s troubling to leave, “Will our community even see this important story?” in the hands of opaque algorithms from social media platforms that were not created with journalism or the public interest in mind; ones that can — and do! — change at any moment.
And both business and editorial minds at publications are aware of a new user base: First-time readers who encounter the publication via one story shared by a social connection. If a publication’s dedicated, paying readership is more precarious, its casual, periodic, and one-time readership is theoretically much broader. But (other than a possible few entering a pay-for-content deal with Facebook), most publications have not yet figured out how to monetize this new casual readership.
Many of the anxieties that publishers and editors express about the current system of news dissemination boil down to: How do we take back control over who gets to see our stories, when they see it, and how?
What this question misses is what the “end users” themselves want — in this case, end users being the people who read, watch, listen to, and share stories. Fifty-five percent of U.S. adults get their news from social media “often” or “sometimes” — suggesting that readers are finding something useful about social media’s content delivery system. So what can we learn about how readers want to encounter stories?
To systems designers, a more interesting version of the question asked by publishers and editors would be: How are users behaving in this system right now, and what does the things they do tell us about what they want and need?
Systems designers focus on human behavior within a complex system. They keep things “objective,” asking only: How are users behaving? Why? And what might this behavior tell us about their priorities?
These designers begin with identifying the components in a system, then mapping the flows — how a key form of data moves through a system. By doing this, designers can identify opportunities for system-wide intervention or innovation.
When we do this, we will often see gaps or bottlenecks in surprising places — and with them, compelling opportunities for change in the future.
Mapping the Ecosystem of News Dissemination
If we were to map out the ecosystem of news dissemination, we could approach it like this:
1. Choose a frame
A map of a complex system will always be incomplete. (We can’t map to infinite granularity — systems maps are often necessarily good-faith approximations.) By placing a frame on a system — What do we most want to understand? — we can draw boundaries around what is most interesting to us.
We’re looking at news dissemination, but what we really want to know is how readers behave — how they find and share stories. So, our systems frame might look like this: How does a person encounter a news story?
2. Map the system objectively, as it is today
A frame will communicate priority, but what is on a map is objective. We’ll put aside what we wish to be true (or wish were not true) — the components and behaviors that go on our map must be true right now.
To be able to make visual sense of the map, we’ll create a key for ourselves.
Circles = people interacting with the story (writers, editors, partners, other outlets, users)
Squares = artifacts containing the story (content types, platforms)
Arrows = the movement of the story (from one person or place to another)
It will take a few tries, but eventually our map might look something like this:
3. Identify patterns… and anomalies
Looking at our map, we can concretely see that users:
- Encounter media stories, rather than seek them out
- Get stories primarily from people they follow, or curated trends in their social media
- Interact publicly with story content — sharing links, screenshots, quotes, and driving more shares of the story
- See individual stories context-free, even if these stories are branded as part of a publication, issue, or series; and
- Don’t pay for their stories.
Interestingly, podcasts and newsletters tell a different story. With podcasts, users:
- Seek out new episodes
- Sample and choose podcasts by trial & error
- Interact publicly with podcast recommendations and reactions
- Hear stories in context, as part of a series; and
- Don’t pay for their podcasts.
With writer newsletters, users:
- Seek out individual writers with whom they’re already familiar, or from people they follow
- Sample, subscribe, and down-select newsletters by trial and error
- Open content directly from their inbox; and
- May pay for their newsletter.
This is where design interpretation and ideation begins. It’s where we go from observers of what is happening to having opinions about why — and, therefore, what should happen. Our systems map suggests at least two compelling innovation paths ahead for how users might encounter news stories.
Innovation Path 1: Ignore the Aggregator
A first path is a version of the editors’ question, above, but draws on the specific, anomalous user behavior with podcasts: How might we create a system in which users have no use for a third-party aggregator between publication and user? Outlets looking to starve aggregators in favor of direct subscription content will need to pay attention to where people don’t rely on an aggregator: namely, podcasts and newsletters. They’ll also need to think boldly — and may find their cues from other entertainment trends.
Like TV streaming services, users seek out podcasts at specific times, often as entertaining filler: during commutes, while cooking, while doing chores, etc. Though they may not finish an episode every time they hit play, users usually listen to podcasts in a series, and in some sort of order. Both podcasts and writer newsletters are easily bingeable, and users quickly develop trust in the subject matter and the hosts/authors over time.
Also like TV streaming services, podcast listeners are aware of the recommendations of others, and sometimes explore new podcasts via platform suggestions, but users’ listening experience is rarely social, and the content is not easily shareable. This makes it much harder for aggregators to “skim” and share content, and means distribution happens almost entirely apart from aggregators.
Could publications take cues from TV streaming services, and incorporate existing behavior from newsletter subscribers and podcast listeners, to revamp how they produce news content?
To build a platform that avoids any need for an aggregator will require something that people are primed to seek out (in the forms, platforms, and times they are already using); that is easy to sample and select; is binge-able; provides context; and is delivered exclusively, in a way that isn’t easily shared on social media.
Innovation Path 2: Empower Users to Control the Aggregator
A second path reflects systemic user reliance on algorithms, but centers users’ needs, wants, and preferences: How might we create a system in which the users have full control over the stories they see?
Designers looking not to starve aggregators but to improve them — by placing power and personalized control in the hands of users — may find cues from concierge-driven brands.
By choosing who to follow, people indirectly curate which stories they will encounter. Whether conscious or not, people place enormous trust in these digital relationships to show them the things they need and want to see.
But this becomes increasingly challenging as users follow more people & more outlets; algorithms curate timelines keyed to clicks and engagement; and other users jam trending topics.
Aggregators dominate dissemination because people use them. Both subscribers and casual users consume news passively: Waiting for a push notifcation, relying on trending, or idly scrolling the newsfeed for an interesting share. To give users more power as their own curators will require something that takes minimal effort; feels personalized; feels transparent (i.e., reflects conscious user intent); and still provides curation services.
These ideas might be not good — they might suck! They are definitely not the only ideas that could speak to the the current system of story dissemination. But by taking a user-centered systems approach — How does a user encounter a story, right now?— we are able to focus our design ideas that bridge existing user behavior with potential for impact and behavior change.
Right now, journalism as an industry is in near-sighted reaction mode: Playing catchup with social media algorithms, and trying to stop the bleeding. By zooming out to a systems lens, we can look at the news ecosystem in context, and start to identify where else interesting things are happening. And that can lead us forward.