Can newsrooms become social platforms?

I’m answering your questions about the future of news technology, and about how news works behind the scenes. If you have a question for me, enter it here, and I’ll answer it in a future post.

Kamran Pakseresht asks:

More and more it seems that stories “develop” in the comment sections. That people often want to understand the reactions to stories just as much as the stories themselves, and sometimes this discovery phase uncovers parts of the story people didn’t even know existed. This pattern seemingly developed alongside crowd sourced services and platforms that utilized the network effects of the internet.

As AI gains prominence, many of these models are trained on the data from these platforms, like Stack Overflow, Quora, Wikipedia, Reddit, traditional forums etc. AI overviews in search engines risk knee-capping the value in these networked models, with users less and less likely to be making it past the overviews provided by Google and now even DuckDuckGo.

It seems newsrooms have never really developed communities around their stories themselves and have always relied on external providers to get this flywheel spinning. Do you see a future where newsrooms promote and contribute to these type of discussions in a more substantive way (understanding that the audience is now part of the story) and potentially becoming a platform in and of themselves? Is the networked model still thriving and can it live side by side with the access patterns being developed around AI?

There’s a lot to unpack here. I’ll break this down into the following questions:

• Will newsrooms embrace audiences as part of the story?
• Can newsrooms become social platforms themselves?
• Can social platform models survive AI?

These questions are particularly interesting to me. Beyond leading newsroom technology at places like ProPublica and The 19th, I spent decades building open source social networking systems that organizations could white label and incorporate into their own communities and brands. Elgg was used by governments, Ivy League universities, social movements, and NGOs. And I’m active in the communities behind the new generation of protocol-driven social media platforms like Bluesky and Mastodon, in part as a board member of A New Social.

Let’s dig into it.

Will newsrooms embrace audiences as part of the story?

Bluntly, the ones that survive will.

I’ve long pushed for newsrooms to change their terminology: “audience” suggests a one-way, broadcast relationship, but relationships on the internet are bidirectional and multi-modal. What every newsroom has is a community, whether it embraces it or not.

Today, newsrooms often address these relationships with two distinct teams:

• Audience: focuses on distribution by posting on social media (including creating video for platforms like TikTok), optimizing headlines, working on SEO, and managing publishing to off-site destinations like Apple News and MSN. The audience team is the primary user for web and social media analytics, analyzing trends to adjust their own posting strategy and reporting back to the newsroom.
• Engagement: focuses on listening to the audience to drive the story-gathering process. The engagement team solicits tips, encourages conversation in the comments, and builds trustworthy relationships that might lead to more and better stories in the future.

In one sense, this means that newsrooms are already embracing their communities. The reader growth driven by audience teams wouldn’t be possible without community, and engagement teams do lead to new reporting. But there’s clearly room for deeper connections here. It’s rare for a newsroom to know who its communities actually are, for one thing, which makes it hard to reach people where they’re at. It’s a little reductive, but you could summarize the role of one team as being the mouth (driving traffic, increasing reach), and another as being the ears (soliciting leads). There’s much to gain by connecting them and having a joined-up community management strategy that seeks to build lasting relationships and learn from a newsroom’s community at least as much as it informs them.

Both teams are traditionally part of the editorial rather than the product function in a newsroom. This also creates a disconnect with product user research: any in-house social platform innovation would need to be a joint venture between the audience, engagement, and product teams. The effort involved would also need to have a clear return on investment that relates to the goals for all three teams.

This setup makes it easy to use incumbent social media platforms like X, Facebook, and Instagram: networks with established toolsets, analytics, and strategies. Only the most experimental teams will take the leap into newer entrants like Bluesky and Mastodon — many added Bluesky to their line-up only when analytics made it very clear that it would increase their audience team goals of increasing click-throughs. And very few organizations will be able to build their own social platforms to serve deeper goals.

Of course, traditional social media referrals are generally declining (particularly on sites like X), search engine referrals are also generally declining (thanks to AI-driven user patterns), and experimentation is badly needed.

Can newsrooms become social platforms themselves?

I think it’s a good idea. Because newsrooms have communities, not audiences, building spaces for community to thrive gives newsrooms first-party data about who is actually engaging with their work and opportunities to build deeper relationships. Doing this on third-party social media platforms might be important for growth and discovery, but it makes a newsroom’s entire community strategy at the mercy of someone else’s business decisions. As we saw with Elon Musk’s acquisition of Twitter and his subsequent move to make it a brand-unsafe, heavily right-wing platform that users are leaving in droves, this can put them in precarious positions.

Many newsrooms see newsletters as a large part of the solution. A newsletter subscription is a direct relationship between the newsroom and a member of the community; the newsroom is more likely to know who that person is, and the person is more likely to receive the newsroom’s content. But it’s still a broadcast model: it’s rare that someone will reply to a newsletter. AI is increasingly creating an intermediation layer for email, too, which means the content might not reach the reader. And privacy tools mean the newsroom might not know who the subscriber is, either.

A lot of people are looking at Vox Media. In August, Vox relaunched its popular SB Nation sports sites to include a community-driven section called “The Feed”. This strategy is explicitly a way to hedge against the decline in search traffic in the wake of AI; it’s not a secret that if it’s successful, it will influence the development of other Vox platforms. The Verge’s editor Nilay Patel has been particularly vocal about the need for social platform integration — and support for next-generation social media protocols, which is a strong hint about where these Vox social platforms are going.

Meanwhile, while smaller startup newsrooms have fewer resources, their team structures are more malleable and they have more to gain by innovating. Many of them use Ghost as their CMS and newsletter platform and embraced its support of the fediverse. For example, articles from 404 Media are fediverse-native and can be directly commented on and reshared by anyone on a compatible platform like Mastodon and Threads.

As a whole, newsrooms tend to be late adopters rather than early innovators. Often, they will wait to see how a technology or approach has been successful elsewhere before trying it themselves (“what is the Times doing?”). So a lot depends on whether these existing initiatives are successful: if they are, we can expect to see a lot more of them. If they aren’t immediately successful, the larger newsrooms are unlikely to take the leap; it’ll take another small, startup newsroom to try something new and make it work.

As we’ve seen with how Ghost’s adoption of fediverse features has translated to use by newsrooms, open source platforms can really help here. It’s rare for a newsroom to build its own CMS; most use WordPress, with a long tail of other platforms bringing up the rear. If WordPress adopted more community features, more newsrooms would use them. A third-party community platform that integrates with existing CMS solutions would go a long way towards helping newsrooms adopt this strategy. Adopting the Automattic WordPress approach would make sense here: the combination of VIP hosted services with an open source codebase gives newsrooms the comfort of not having to maintain their own infrastructure, safe in the knowledge that they have a credible exit if the service fails them in some way.

There’s one other wrinkle: Mississippi. The state now has a law on the books that requires users to verify their ages before using social media sites. This has led Bluesky to block use in the state entirely. If newsrooms became social platforms, they would have a choice: either follow in Bluesky’s footsteps or ask for verification for users in Mississippi up-front. This is clear friction for the kind of growth audience teams are looking for, and creates extra legal liability at a time when newsrooms are politically under attack. So it might well be that American newsrooms are shy to create these kinds of features in the current environment.

Can social platform models survive AI?

Can they? Sure. Will they? Maybe.

It’s isn’t fully one yet, but AI has the potential to bring about a paradigm shift in terms of the way people access information. Previous paradigm shifts have included the web itself, the iPhone, and the rise of near-ubiquitous connectivity. In each case, the organizations that embraced change and dared to change their strategies in the face of trends survived in far greater numbers than the organizations that clung to their existing strategies for longer.

The way to do this is to go back to the users, understand who they are and what they need, and build something that addresses their needs well using ongoing trends as a guide. If you build for now, you’ll always be building for the past: it takes time to establish any kind of new product or strategy, and the world will have changed in the meantime. So you need to make an informed bet on the future — and the only way to do that is to deeply understand where the world is going, and to deeply understand the people you’re trying to reach.

AI summaries on sites like Google are designed to answer definitive questions. But neither journalism nor communities map well to a question-and-answer model. The likes of Stack Overflow and Quora might have to adapt a little, but the real value is in qualitative conversation and contextual information, not in a deterministic answer. For journalism, that means doubling down on providing stories and context you can’t find anywhere else; for communities, that means promoting really great, substantive discussion. In his question, Kamran mentioned that “stories develop in the comment sections”; AI can’t hope to replace that. And in fact, AI could help users find communities and conversations that really provide value to the user.

What we don’t know yet is what future AI-powered interaction models might look like. It won’t be a chatbot, just like most of us aren’t using MS-DOS or UNIX. AI agents might be a big part of new browsers, creating new interfaces and summaries of publisher-provided information and community conversations. In that world, advertising-supported sites are dead, because the ads will never be displayed, and it may be harder to build direct relationships with users.

While it certainly requires adaptation, it’s actually a progression of an ongoing trend. Browsers were always meant to be extensible to modify content on the page: that’s why extensions exist. Users were always meant to be able to modify CSS for themselves to make sites more readable. And most modern browsers ship with ad blockers.

I’m not going to cry for the demise of ads. I do think modern AI vendors have an exploitative relationship with publishers; for example, the Anthropic class action settlement indicates that author rights have been infringed through its storing of pirated material for use as training data. Setting up AI vendors as a new generation of information gatekeepers would be bad for the world.

But my ideological feelings about it don’t have a bearing on whether it will actually happen or not, and newsrooms and community platforms both have to be ready for a future where it will. AI might disappear at the end of its hype cycle like blockchain has; if it does become the future of computing, organizations like Mozilla that are pushing for more ethical models might have enough influence to make the ecosystem more pro-social. Regulation from places like Europe may place important restrictions on the industry’s worst impulses. We just don’t know yet.

I think we can be reasonably confident that it won’t look like Meta’s bizarre experiments making the dead internet theory even more of a reality. OpenAI is building a social network and I don’t have high hopes that this will point to the future of social communication either. The social future will likely be established by a new startup with nothing to lose, as has been the case with just about every paradigm shift. It’s very possible that Bluesky, or more specifically its underlying AT protocol, will be a part of this future. But there’s everything to play for.

So what are the implications?

Regardless of what the future holds, newsrooms have a lot to gain by deepening their understanding of, and engagement with, their communities. They need to know who they are as people and meet them where they’re at. That likely means deepening their product efforts around community and getting comfortable with, if not becoming platforms themselves, integrating more deeply with social media protocols.

I would like to see three things:

• A remodel of audience and engagement teams around a more holistic community strategy.
• More product experimentation in the model of Vox’s SB Nation redesign.
• Community platform vendors that embrace integration with CMS solutions in order to make this experimentation easier.

Regardless of the future of computing, these moves create more trustworthy, enduring, direct relationships. Ultimately, what drives a newsroom isn’t the underlying technology; a platform can only provide tools to support community and relationships. What really matters is the people. Newsrooms have to understand that they’re in a relationship business, and reorganize themselves around that simple fact.

Do you have a question about newsroom technology? Ask it here and I may answer it in a future post.