Expanding from print into digital channels gave the media sector an important lesson in scaling. The race to figure out AI for publishers makes this a good time to revisit it.
A generation ago, for example, newspapers and magazines had to lay out all their content within a limited number of pages. The internet’s arrival changed that by allowing editors to publish far more stories across their websites, social media accounts, mobile apps, email newsletters, and more.
For the first time, journalism could scale so that media organizations could give a story as much coverage as they wanted. The trade-off was that doing so across digital channels other than their websites risked high costs and loss of control.
AI creates a different opportunity to scale across multiple parts of the business, such as content creation, audience development, ad sales, and product development. The opportunity AI presents makes it worthy of becoming a bigger budget line item, but only if media orgs are ready to capture ROI. Determining that will take some investigative reporting.
Spending on AI isn’t the same as being ready for it
According to market research firm Gartner Inc., more than 15% of marketing budgets go to AI, but just 30% say they’ve achieved mature readiness, or are fully prepared to treat AI as an operating system versus a new set of tools.
AI readiness in publishing specifically is more difficult to benchmark, but the industry’s intentions are clear. A WAN-INFRA’s World Press Survey found more than 87% of media companies are planning to invest in AI and automation over the next twelve months. A quarter cited it as their highest spending priority.
Pouring money into AI needs to be coupled with a clear sense of where it should deliver value, and this looks different in media than it does in other sectors. As the technology gets piloted and deployed, it should be helping:
- Editors and reporters produce more variations on the stories they write, including localized versions and alternate formats, without degrading quality or accuracy.
- Audience development teams see which AI referrals lead to signups and paid subscriptions, as well as opportunities to personalize experiences to drive loyalty.
- Product teams create and ship additional newsletters and sites while maintaining consistently high performance and aligning with governance requirements.
If this sounds like a tall order, it’s no easier outside of publishing. According to WordPress VIP and Human Made’s AI Readiness Report, 77% of teams use AI, but only 57% say they’re strategically aligned. Even more challenging, CMS and workflow constraints limit progress for 6 in 10 organizations.
The best way to avoid those pitfalls and ensure you can scale AI for publishers is to plan with more than your flagship site in mind.
Why multisite environments are an ideal AI readiness test
Take every content operations problem publishers tend to have (such as manual tagging, review and approval breakdowns, and limitations in reusing structured content) and multiply them across a dozen websites or more.
When you’re running regional editions, vertical publications, and event hubs, it’s not unrealistic to have small problems cascade across an entire multisite environment.
AI can help automate solutions to many multisite issues, and it’s also a great way to get a fuller understanding of all your different content types, workflows, audience expectations, and monetization journeys.
Instead of running an AI pilot across your entire organization, a multisite environment lets you test the technology within a single site or group of sites. This makes it easier to measure the results and reverse decisions if need be.
Running a multisite environment on a single CMS puts you in an even better position to adopt AI at scale because it allows shared governance based on consistent, structured data and one content model. Take Recurrent Ventures, which pulled off seven site migrations in eight months to a single, block-based platform.
Once you’ve started using your multisite environment as an AI readiness test bed:
1. Train the newsroom to work on AI
No matter which site their work appears on, journalists should understand how they can easily adapt platforms and tools to enhance what they’re already doing. Codify where AI can be used for brainstorming, researching, and assisting in actual writing. Explain where AI could accelerate review and approval processes and where editors need to be directly involved before a story goes live.
It helps when you’re already running on an AI CMS designed to make editorial teams more productive. For instance, New Scientist found moving to WordPress VIP means images load instantly versus waiting for 10 minutes or more, while block-based editing reusable patterns reduced the time spent hunting through menus, HTML, and old content.
2. Train audience dev teams to measure what AI changes
AI should be able to help fine-tune audience segmentation and testing to achieve greater precision and deliver deeper insights. That needs to be balanced by clear guidelines and communication to site visitors about how their data is used. Being able to audit and explain AI outputs will be critical to prevent actions that might damage an audience’s trust.
Start by implementing an analytics platform that can handle a multisite environment. USA TODAY Co. did that by adopting Parse.ly, which is built into WordPress VIP and now provides insight on content performance across its more than 600 sites.
3. Train product teams on how to assess and explore AI use cases
AI in journalism is going to morph significantly based on how audiences change the way they seek out information, how quickly editors and reporters improve their skills, and how well websites meet both groups’ needs. A multisite environment gives product development professionals in media the best of both worlds: a centralized set of integrations and APIs with the ability to customize for individual brands.
AI tools and platforms should be assessed based on their ability to improve discovery on answer engines, ease localization and translation, and reduce security and governance risks.
Although a lot of AI activity is already happening in channels like social media and search tools, WordPress CTO Brian Alvey argues that websites (and newsletters that distribute content published on them) offer media organizations the greatest control and direct relationship with audiences.
“Compared to everything else, social and LLMs — those are very powerful middlemen standing in front of you and your audience,” Brian said during an episode of the Press Gazette’s Future of Media Explained podcast. “They’re never going to be as good (as your websites).”
Making the move towards AI in multisite media organizations
Publishers should begin their AI journey with the end in mind: where any of the digital titles in their portfolio will be easily discovered in search, where audience teams can track referrals to conversion, and where AI agent access is governed by strictly enforced policies.
Multisite environments will teach publishers quickly what kinds of AI deployments work, what don’t, and where humans need to be an integral part of business processes.
Simple pilot tests could include automating metadata with AI or localizing headlines for a regional edition. Depending on the results, stakeholders can come together to determine quality standards, the degree of human oversight or review, and what kind of disclosures they need to offer audiences.
As those pilots turn into a scaled deployment, everything AI does should ladder back to established media metrics such as engaged time per visitor, recirculation rate, and opt-in for newsletters and new content experiences.
Having a CMS that lets you manage multiple sites is foundational in this approach, as is a CMS that offers headless and hybrid architectures, easy integration, strong performance and security, and support to help adapt to business demands that will emerge later on.
Much like bringing journalism to digital channels, AI for publishers is going to become a natural evolution of how the industry operates. It’s almost like one of those news stories that continues to get coverage over an extended period. At press time, “AI readiness” is the headline on that story.
