01. Introduction
For many digital publishers, Google Discover has become one of the most exciting and frustrating sources of traffic at the same time.
A story can suddenly reach hundreds of thousands of readers. A publisher can experience a remarkable traffic spike without the user ever typing a search query. Then, just as quickly, that visibility can disappear.
This has naturally made Discover an important part of the growth strategy for news and sports websites. Publishers invest in stronger headlines, better imagery, author credibility, technical performance and content strategies designed partly around the possibility of appearing in Discover.
There is nothing wrong with that. Discover can be enormously valuable.
The problem begins when Discover becomes the strategy rather than one part of it.
The digital environment around publishing is changing too quickly for any serious media organisation to build its future around a single external distribution channel. Search itself is changing. AI platforms are becoming discovery environments. Social platforms continue to influence how stories travel. Direct audiences remain difficult to build but increasingly valuable.
The question publishers should be asking, therefore, is not simply: How do we get more traffic from Google Discover?
It is: How do we build a publishing platform that can grow even when the way people discover news changes?
02. Discover traffic is valuable, but it is borrowed traffic
The attraction of Google Discover is obvious. Unlike traditional search, users do not necessarily need to express an intention. Google recommends content based on interests, behaviour and context, giving publishers access to audiences that may never have actively searched for their stories.
For news publishers, the scale can be significant.
But there is an important distinction between attracting an audience and owning a relationship with that audience.
A visitor arriving through Discover is still arriving because another platform decided to recommend the content. The publisher does not control when that recommendation happens, how frequently it happens or how much exposure a particular story receives.
This is not unique to Google. The same principle applies to Facebook, Instagram, TikTok and practically every external distribution platform.
Algorithms change. Interfaces change. User behaviour changes. Features that generate enormous reach today may become less important tomorrow.
A sustainable publishing strategy therefore cannot be based on predicting what an algorithm wants. It has to be based on building something valuable enough to perform across multiple environments.
03. The website still matters more than it appears
With so much discussion around platforms, feeds and AI, it is tempting to think of the publisher’s website simply as the destination where the article happens to live.
That significantly underestimates its role.
The website is the part of the publishing ecosystem that the publisher actually controls. Its architecture determines how content is organised, how quickly pages load, how easily search engines crawl them, how stories connect with topics and authors, and whether a casual visitor has any reason to continue reading after finishing the first article.
A modern news platform should not be designed purely around publishing individual URLs. It should represent the structure of the publication itself.
Authors should be meaningful entities. Topics should connect related coverage. Sports competitions, teams, people, locations and events should have useful destinations rather than existing only as loosely connected tags.
This is one reason why the architecture of a modern news portal has become inseparable from SEO, Google Discover and AI visibility. The same structural decisions increasingly affect all three.
A publisher that gets the foundations right is not optimising for one algorithm. It is creating digital infrastructure that can support multiple forms of discovery.
04. Search is becoming much bigger than ten blue links
Traditional SEO is not disappearing, but the environment around it is changing.
A user looking for information may still search on Google and visit a publisher directly. Increasingly, however, the journey can include AI Overviews, conversational search, ChatGPT, Gemini and other AI driven interfaces that attempt to answer the question before the user visits a website.
For publishers, this creates an uncomfortable reality. Producing good content is no longer enough to guarantee that the relationship between the content and the audience will happen on the publisher’s own platform.
That makes clarity, authority and structure increasingly important.
A machine needs to understand who wrote an article, when it was published, what it is about, which organisations or people it references and how it relates to the publisher’s wider coverage.
This is where the conversation moves beyond traditional keyword based SEO. Good search engine optimization remains fundamental, but publishers also need to think about entities, structured data, authorship, topical authority and the consistency of information across their digital presence.
In other words, publishers should not optimise only for rankings. They should optimise to be understood.
05. AI visibility is becoming part of publishing strategy
This leads to another question that publishers increasingly need to address: what happens when the reader asks an AI system about a story rather than searching for it?
Imagine a user asking for the latest developments around a football club, an election, a major court case or an economic story. The answer may be assembled from several sources without the traditional search journey ever taking place.
Being represented accurately in those environments will matter.
There is still a great deal of uncertainty around how generative discovery will develop, and publishers should be cautious about anyone promising a formula that guarantees AI citations. But the direction is important enough to take seriously.
Clear authorship, strong editorial expertise, well structured content, consistent entities and useful original reporting all strengthen the information foundation from which search and AI systems operate.
This is also why Generative Engine Optimization should not be treated as a replacement for SEO. The two increasingly overlap. Both depend on producing credible information and making its meaning understandable to machines.
The objective is not to write articles for robots. It is to remove ambiguity around content that was written for people.
06. Publishers need direct relationships again
There is another side to this discussion that has very little to do with algorithms.
For years, publishers have invested enormous energy in acquiring audiences while often doing relatively little to retain them.
A reader arrives from Google, reads an article and leaves. Another arrives through Facebook, reads an article and leaves. A third comes from Discover and does exactly the same thing.
Traffic increases, but the relationship does not necessarily become stronger.
This is where direct channels become strategically important again.
A good newsletter can create a recurring relationship with readers. Push notifications can bring users back for genuinely important developments. Personalisation can help readers follow the topics that matter to them. Registration can allow the experience to become more relevant over time.
And for publishers with audiences that return frequently, a well designed mobile application can become much more than another place to display the same articles.
It can create habits.
That distinction matters. Traffic is something a publisher receives. An audience is something a publisher builds.
07. A news app should not simply be the website in an app
Many publishers made an understandable mistake during the first generation of news applications. They essentially placed the website inside an app and expected users to download it.
There was often little reason to do so.
A modern publishing app needs to provide additional value. That might mean personalised feeds, saved stories, following specific journalists or teams, live scores, real time updates, intelligent notifications or a better experience around live events.
The opportunity is particularly interesting for sports publishers. A reader may arrive because of one article about a football team, but an app can build a much broader relationship around that interest through fixtures, standings, live coverage, statistics and personalised alerts.
This is fundamentally different from chasing another pageview.
It moves the publisher from individual content consumption towards a recurring product experience.
08. Publishers also need to think beyond the article
The article remains the core unit of digital journalism, but it does not have to be the only one.
A sports publisher may have structured data around teams, competitions, fixtures and standings. A financial publisher may have companies, markets and economic indicators. A general news organisation may build rich topic pages around people, organisations, places and continuing stories.
When this information is properly structured, it can support much more than navigation.
It can improve related content, internal search, recommendation systems, newsletters, apps, APIs and eventually AI powered experiences.
This is why metadata is becoming a competitive advantage. As Pavla’s existing analysis notes, for a news article metadata can describe its author, publication date, topic, organisations mentioned and related events, giving systems the context needed to connect and reuse information effectively.
For publishers producing hundreds or thousands of stories every month, these relationships accumulate into something much more valuable than an archive of old URLs.
They become a knowledge base.
09. Performance and infrastructure are editorial issues too
None of this works particularly well if the underlying publishing platform cannot cope with the realities of news.
News traffic is unusual because demand can change in minutes. A breaking story, major sporting event or unexpected development can send tens of thousands of readers to a site almost simultaneously.
At that moment, infrastructure stops being an IT department issue and becomes a publishing issue.
A slow or unavailable website loses readers precisely when the publication is most relevant. It can also affect advertising delivery, subscription journeys and the overall perception of the brand.
The same applies to everyday performance. Readers do not care why a page is slow. They simply experience the friction.
Publishers therefore need infrastructure designed around traffic spikes, caching, efficient content delivery, image optimisation, database performance and resilience. Good hosting architecture is not simply about keeping a server online. For a high traffic publisher, it is part of the product.
10. The goal is not independence from platforms
Thinking beyond Discover does not mean abandoning Discover.
Nor does it mean abandoning Google Search, social media or any other platform capable of introducing journalism to new audiences. These channels remain extremely valuable and publishers should continue to understand how they work.
The mistake is dependence.
A stronger model uses external platforms for discovery while simultaneously investing in the assets the publisher controls: its technology, content architecture, brand, journalists, first party audience, newsletters, apps and data.
When Discover performs exceptionally well, the publisher benefits. When search behaviour changes, the organisation can adapt. When a new AI discovery environment emerges, its content is structured well enough to participate. When social algorithms change again, there is still a direct audience to return to.
That is a fundamentally more resilient position.
11. From traffic strategy to audience strategy
Perhaps the most important change is conceptual.
Publishers have spent years discussing traffic acquisition. The next phase should be about audience development.
Those two ideas sound similar, but they lead to very different decisions.
A traffic strategy asks how to generate another million pageviews. An audience strategy asks how many of those readers will recognise the publication tomorrow, return voluntarily, subscribe to a newsletter, install the app, follow a journalist or choose the same source when another major story breaks.
Google Discover can play an important role in that journey. In some cases, an enormous one.
But it should be the beginning of the relationship, not the relationship itself.
The publishers best positioned for what comes next will not necessarily be those that discover the perfect formula for every algorithmic change. They will be the ones that build strong journalism on top of equally strong digital infrastructure, make their information understandable across search and AI environments, and give readers enough value to come back without waiting for an algorithm to send them.
Discover can deliver remarkable traffic. Building an audience is what turns that traffic into a publishing business.
