Over the past five editions of the Samhub Signal, I have covered first-party data as the publisher's revenue moat, what personalization actually means in 2026, and what a complete first-party data stack costs when you map it honestly. 

The thread running through all of it has been the same: publishers who control their first-party audience data have options. Publishers who do not are at the mercy of whoever controls the algorithm this quarter.

But data infrastructure is a prerequisite, not a destination. And the question I hear most often when I sit down with publishing executives right now is not how to build the stack. It is what to do once you have it. 

Specifically: how do you convert what you know about your audience into revenue streams that do not all compress at the same time?

That is what this edition is about. I am calling it the revenue stack, the deliberate construction of multiple income sources, each drawing on the same underlying audience intelligence, each insulating the others when one channel contracts. 

It is not a new idea. But the publishers who are actually executing it in 2026 are doing it differently from how most of the industry still thinks about diversification.

The problem with how publishers talk about diversification

Most conversations about publisher revenue diversification start in the wrong place. They start with a list of revenue types, subscriptions, events, commerce, licensing, branded content, and they treat diversification as the act of adding more lines to that list. 

Find a new channel. Launch a product. Run some events.

The publishers I see doing this tend to end up with a collection of small bets, none of which are large enough to matter, all of which require operational overhead they did not budget for. 

The commerce vertical underperforms because the editorial team does not trust it. The events team runs one conference and then burns out. The licensing conversations go nowhere because there is no clear story about what the audience is worth.

The publishers who are actually diversifying well are not adding revenue lines. They are building revenue architecture. The difference is that architecture starts from a question: what do we know about our audience that somebody else would pay for, or that our readers would pay for directly? Everything flows from that.

If you cannot answer that question with specifics, not demographic approximations, but actual signals about intent, interest, and behaviour, then most of the revenue streams that get described as diversification are not really available to you. 

You are adding products to a foundation that cannot support them.

Reader revenue: the floor you have to build before you can build anything else

Subscriptions and memberships get talked about as a growth strategy. They are better understood as a test of editorial clarity. 

If a reader will not pay for your content, the problem is rarely the paywall. It is that the content does not do something specific enough to justify payment, it covers the same ground as a dozen free alternatives, written for a general audience rather than the reader who most needs what you specifically produce.

The publishers making subscriptions work in 2026 have all done the same uncomfortable exercise: they have identified the two or three things their audience genuinely cannot get anywhere else and built the paid product around those things. 

Not their best content in a general sense, but the content that creates the kind of dependency where missing it costs the reader something real. Data, access, analysis that cannot be replicated by an AI summary of the public record.

What subscriptions do that no other revenue stream does is generate consented, identified, first-party data at scale. A paid subscriber has raised their hand. They have told you something about what they value, they have given you a payment relationship, and in most cases they have consented to a level of data use that anonymous traffic cannot support. 

That data layer is the foundation for direct ad sales, for events, for licensing conversations - for every other revenue stream that depends on being able to say something credible about who your audience is.

The Financial Times understood this before most publishers were willing to admit it. Their subscription pivot in the early 2010s was not primarily about subscription revenue. 

It was about building an identity graph on top of their most engaged readers, one that would let them sell advertising at rates that programmatic could not touch, because they could prove the audience rather than approximate it.

Events: monetising the trust gap

I have a simple mental model for why events work as a publisher revenue stream. Readers who follow you regularly are predisposed to trust the speakers you curate, the conversations you host, the sponsors you bring into the room. That trust transfers. And it is worth significantly more than a CPM.

The Financial Times generates more from events than from print advertising. The Economist runs an events business that effectively cross-subsidises editorial. These are not side projects that happened to work, they are the result of publishers recognising that their audience wants to be in the room, not just read about what happened in it.

What makes events particularly valuable in the current environment is that they monetise something platforms cannot replicate: physical or intimate digital access to an audience that has been curated by editorial judgement. 

A CMO who attends a FT live event is not there because of a targeting algorithm. They are there because they trust the brand, and because the other people in the room reflect a standard of curation they cannot get from a LinkedIn conference.

The mistake I see most often is publishers trying to launch events at a scale they cannot sustain. The better entry point is smaller than you think. A dinner for forty people, carefully invited, with a genuinely interesting conversation and two or three relevant sponsors, will teach you more about how your audience monetises than a conference of three hundred that costs twice as much to produce and generates half the revenue per attendee. 

Start there. Build the case. Then scale it.

AI licensing: the open revenue window

I want to be honest about the uncertainty here, because I think the conversations I have seen in the trade press have ranged from too enthusiastic to too dismissive, and neither is useful.

What is true: AI companies need high-quality, trustworthy, human-authored text to train and ground their models. Publishers who have spent decades producing it are sitting on an asset that has a new category of buyer. 

The Associated Press, Axel Springer, The Atlantic, News Corp, the list of publishers who have signed licensing agreements with AI labs has grown substantially over the past eighteen months, and the terms, while not always disclosed, are significant enough that several publishers have cited them as material to their financial position.

However, when I was at the INMA event in Berlin earlier this year I heard several high level officials from media houses all over the world say this: Close the door for AI, don’t let them crawl your content, don’t let them in the door until you are sure that there is something in it for you!

What is also true: the window is not permanent. The AI companies that are paying for content today are paying partly because they need it and partly because they want to establish that they are willing to pay, ahead of regulatory frameworks that will eventually require them to. 

Publishers who wait for the regulatory outcome may find that the voluntary licensing market has already been shaped by the publishers who moved first.

The strategic question is not whether to license. It is at what price, for what use cases, and with what restrictions on how your content can be used in AI-generated outputs that compete with your own. 

These are not simple questions, and they require legal and commercial sophistication that most regional publishers do not have in-house. But the conversations are worth having now rather than later. 

The leverage that comes from being a large enough content holder to negotiate from strength diminishes as more publishers accept lower terms.

First-party data as a product: what your audience intelligence is actually worth

This is the revenue stream that requires the most infrastructure to unlock, and the one with the most asymmetric upside for publishers who get there.

The advertisers who are most aggressively looking for alternatives to Google and Meta right now are not looking for reach. They can buy reach anywhere. 

What they cannot buy from the platforms (at least not transparently), is certainty about who they are reaching and why those people are likely to care about what they are selling. Publishers who can answer that question with first-party data are selling something genuinely scarce.

When COPE (Content Performance Group) presented their first-party data results at INMA Berlin earlier this year, the numbers that stood out were not the CPM uplifts, though a 200% CTR improvement is significant. 

What stood out was that advertisers voluntarily moved performance budgets from Google to COPE's data-driven placements, not because COPE had better reach, but because COPE could tell them something about their audience that Google's black-boxed targeting could not. That is a different kind of selling conversation, and it commands a different kind of margin.

The prerequisite is a clean, consented, usable data layer. Not a third-party approximation of your audience: your audience, identified, segmented by signals that are specific to your content and your reader relationship. 

Most publishers do not have this yet. The ones building it now are operating with an advantage that compounds over time, because the data improves with every interaction and the audience understanding deepens with every campaign. 

In Sweden we see that publishers are connecting with brands through their own unified audience standard and analytics, through platforms such as Samhub, tailored to media.

The agency model: from media provider to partner

There is a shift happening among the regional and specialist publishers I speak to that does not get enough attention, partly because it does not fit neatly into the standard taxonomy of publisher revenue streams. 

It is the move from selling media space to selling media thinking: becoming, in effect, a strategic partner to local and regional advertisers rather than just a channel for their budgets.

The logic is straightforward, even if the execution is not. A regional publisher operating in a specific geography or vertical has something that no global platform can replicate: genuine, accumulated intelligence about a local audience. 

They know which industries dominate the local economy. They know which businesses are growing and which are contracting. They know what their readers care about, what they distrust, what they respond to, and, if they have built the right data infrastructure, they can prove it with first-party signals rather than approximations.

Most local advertisers, particularly the SMEs that make up the bulk of regional advertising budgets, are navigating a media landscape they do not fully understand. They are being sold Google campaigns by local agencies, Meta campaigns as well, and possibly programmatic packages by networks whose incentive is volume rather than outcome. 

What they do not have is someone who understands their local market, knows their audience, and can take a view on where their budget should actually go, including channels the publisher does not own.

This is the opening. Publishers who step into it are not just selling their own inventory. They are offering media planning and buying across multiple channels, with the local audience intelligence as the differentiator. 

They are providing analytics that tell an advertiser not just how their campaign performed on the publisher's properties, but how it performed in the context of the wider media mix. 

They are consulting on creative, on message, on the combination of channels that makes sense for a business trying to reach a specific local audience at a specific moment.

NRC Media in the Netherlands has been building in this direction for several years, developing advertiser services that extend well beyond traditional media sales into creative production, performance consulting, and audience strategy. 

In Sweden, Bonnier News Local has experimented with similar models at the regional level, positioning local editorial brands as market intelligence partners rather than just ad platforms. 

The revenue per advertiser relationship looks fundamentally different when the conversation starts with strategy rather than rate cards.

The commercial model that makes this work is typically a retainer or partnership fee alongside media spend, the consulting component priced separately from the inventory, reflecting the time and expertise involved. 

Some publishers have created dedicated agency units sitting alongside the editorial operation. Others have trained existing sales teams to have a different kind of conversation. Neither approach is simple, and both require a genuine investment in capability: in data tools, in analytical skills, in the confidence to have a conversation about media strategy with a client who previously just received a proposal.

What makes the agency model particularly compelling as part of a diversified revenue stack is that it is structurally resistant to the pressures that compress other streams. CPMs are set by markets. Subscription growth has a ceiling defined by your audience size. 

But the value of genuine local expertise, the ability to help an advertiser understand their market and reach their audience more effectively than a platform algorithm can, does not compress in the same way. It is a relationship business, and relationships compound.

The prerequisite, again, is knowing your audience specifically enough that the advice you give is worth paying for. 

A publisher who can tell a local car dealership not just that they reach 40,000 readers in the region, but who among those readers is in a car-buying window, what price point they are considering, and which channels they use to research, that publisher is having a different conversation than one selling a banner package. 

The data infrastructure makes the agency model credible. Without it, it is just a pitch amongst hundreds of other media agency pitches.

The common thread

Every revenue stream I have described is downstream of the same thing: knowing your audience specifically enough to say something credible about what they value, what they will pay for, and what they respond to.

Reader revenue requires knowing which content creates genuine dependency. Events require knowing which conversations your audience will travel for. AI licensing requires being able to demonstrate the quality and provenance of what you have produced. 

Data products require a first-party infrastructure that reflects real, consented audience behaviour rather than third-party inference. The agency model requires knowing your local market well enough that your advice is worth paying for independently of your inventory.

The publishers I am watching navigate 2026 well are not the ones who are running the hardest. They are the ones who have made the foundational investment, in identity infrastructure, in audience understanding, in the data layer that makes every other revenue stream credible. And are now executing from a position of relative clarity.

The ones who are struggling are still waiting to start that investment, on the grounds that they have more urgent priorities. In my experience, that calculation tends not to resolve in their favour.

If you are at a publisher working through what this looks like in practice: the sequencing, the infrastructure requirements, what is realistic at your scale, I am happy to compare notes. You know where to find me (or have a look at our summer edition newsletter we sent out in July, detailing the infrastructure requirements for any media house).

Until the next edition.

/Martin Bergqvist, CEO & Founder