For its fifth edition, the “Tech Stack” study, conducted by the association of the same name in partnership with the Sciences Po Journalism School, paints the most complete picture to date of the tools that French news publishers rely on to produce, distribute and monetise news. Four tool families, 118 contributions, nearly 5.5 billion monthly page views described: the findings of the 2026 edition are now available to download.

Which tools do French news publishers actually use to produce, distribute and monetise their digital content? Since 2021, the “Tech Stack” survey has answered this question by drawing directly on the voices of the professionals who work with these tools day to day: chief technology officers, newsroom managers, product managers, marketing managers and ad sales house directors. Between in-house solutions and external vendors, they all administer a technical apparatus in constant recomposition. The very word “stack” captures the iterative nature of this accumulation, driven by the evolution of technologies and the transformation of usages.

Launched in 2021 on the initiative of Marion Wyss, an online-media professional and specialist in subscriber-acquisition strategies (and Audiencers’ Publisher), the Tech Stack reaches its fifth edition in 2026. As early as 2022, Sciences Po joined the project to ensure its longevity in the form of an annual update. Its founding principle remains unchanged: to consolidate the full range of practices in order to draw up a complete overview, while guaranteeing each respondent strict confidentiality over the solutions they report using.

A survey of unprecedented scope

The 2026 edition is based on four separate questionnaires: marketing, editorial, advertising and technical tools. Completed by the managers of the relevant teams, each solution cited was analysed through a twofold approach: its market share among respondents on the one hand, and an assessment of its quality on the other. Beyond listing the tools they use, respondents rate them: each professional gives their solution a satisfaction score from 1 to 10 and specifies, in free-text comments, what wins them over or frustrates them in terms of quality and everyday use. It is the aggregation of these scores that makes it possible to measure whether, in France, publishers feel well served by their CMS, their subscription manager or their adserver. And it is this cross-analysis of equipment rate and satisfaction that makes the study a genuine decision-support tool, allowing each publisher to guide its own equipment choices by drawing on the experience of its peers.

Measured against the combined audience of all participating titles, the study describes the Tech Stack underpinning more than five billion monthly page views. Contributors include the majority of national daily press titles, several audiovisual brands, magazine titles, regional and trade press groups, as well as numerous pure players. In keeping with the confidentiality commitment made to respondents, all responses were consolidated and processed anonymously, but the list of respondents is set out in the study’s introduction and in each section.

Publishers now properly equipped

The first takeaway from the 2026 edition lies in the convergence of scores. The four tool families obtain closely aligned averages: 7.1/10 for marketing, 7.2 for editorial, and 7.4 for both advertising and technical tools. The message is clear: French publishers are now properly equipped. No category collapses, and none reaches excellence. The gaps lie at the extremes, from the success of digital advertising audio to the struggles of the subscription manager — the backbone of the subscriber relationship, still largely inherited from the print era. Over three years, the trend is one of slow improvement as stacks stabilise.

This maturity shifts the challenge. Everywhere, value no longer lies in the tool itself but in its ability to connect to others. Marketing calls for a unified data layer; editorial dreams of a CMS turned cockpit, linking planning, publishing, social media and measurement; the technical side seeks observability and orchestration; and advertising, the reconciliation of its fragmented signals. The 2026 stack is not under-equipped: it is under-orchestrated. To this integration tension is added, among legacy press groups, a persistent divide between print-inherited foundations and digital building blocks — a divide that dreaded migrations still struggle to close.

Two underlying forces: AI and sovereignty

Two dynamics run through all the chapters. First, artificial intelligence, now present in almost every use case (nearly nine publishers out of ten on the marketing side) but still confined to assistance and individual productivity. Its industrialisation, plugged into data, is shaping up to be the next front of differentiation. Second, sovereignty, which is emerging as a structuring criterion in the face of dependence on the major US players, from web analytics to the adserver, from cloud to language models, and in the face of rising costs (licences, cloud, tokens) that everyone is seeking to keep under control.

One constant remains, a common thread throughout the study: in-house solutions, when designed together with the teams, consistently earn the best scores. Proof that satisfaction stems less from a tool’s power than from its right fit with needs. That is where the 2027 challenge lies: not in accumulating new building blocks, but in the art of connecting them.

Download the study

The full results — a detailed analysis of the four tool families, equipment rates, satisfaction scores and publisher verbatims — are available in the complete summary of the “Tech Stack 2026” survey.

→ Download the Tech Stack 2026 study (PDF)

A study by the Tech Stack association, in partnership with the Sciences Po Journalism School.

The post The Publishers’ Tech Stack Study: findings from the 2026 survey appeared first on Audiencers.

The Audiencers' team spent the first three days of June in Marseille for WAN-IFRA's 77th World News Media Congress (1–3 June 2026). Three days, one dominant question: what does AI mean for the value, and future of journalism? We sat through a dozen sessions to share a wrap-up of the event, and the conversations that mattered most.

 3 things to take from this recap
> Rethink your "traffic" metric: consider tracking the share of your content AI can't easily copy or summarize, alongside reach
> If you're negotiating an AI licensing deal, know whether you're getting a product partnership or just a defensive payout, they're not the same thing
> Audit how your newsroom's time actually splits across production, pre-, and post-publication before assuming AI tools are the fix

The AI reckoning: rights, traffic, and the first real deals

AG Sulzberger (NYT): the “original sin,” and an 8-point plan

The congress opened on its hardest line. AG Sulzberger (Chairman & Publisher, The New York Times) argued that generative AI is built on an “original sin” (the unprecedented theft of publishers’ intellectual property) and that this theft is now re-enacted “countless times every single day.” His most effective passage broke an AI model into four inputs: talent, compute, energy and data. The first three are paid for, handsomely; the fourth, books, films, music, journalism, is taken without consent or compensation. No tech CEO would ask their engineers to work for free or steal chips from an Nvidia plant; on data, “we just take” has become the norm.

“Tech giants strip-mine news websites… that otherwise would go to the news organisations that created this work.”

AG Sulzberger, Chairman & Publisher, The New York Times

He was careful to add he is not anti-AI (the NYT uses these tools internally) and that his target is the companies’ choices, not the technology.

The scale he put on the table: the NYT publishes ~500,000 original works a year for over $2bn, and remains the single largest source of proprietary data in a major dataset used to train many different models. Meanwhile the Google click is 10x harder than a decade ago, rival AI models send 96% less referral traffic than a Google search, and the largest news sites tracked by comScore have lost 45%+ of traffic in four years. His action plan came in eight points — four to defend your rights (stand up for them, deal carefully, push legislators, join together) and four to strengthen your title (use AI the right way, be a destination first, focus on original reporting, explain why journalism matters). He closed by restoring the half of the Stewart Brand quote everyone forgets:

The post Inside WAN-IFRA Marseille 2026: the deals, the data, and the fight for what journalism is worth appeared first on Audiencers.

This article summarizes the session presented by Maxime Moné (CEO of Poool) and Virginie Clève (founder of Largow, a consulting agency) at the Audiencers Festival Paris on September 16, 2025. Together, they explored the question of “SEO & Paywalls: how can you combine audience and subscriptions in the age of AI?”

This presentation summarized in 5 points: 

- Choosing the paywall type (super hard/hard/soft) should be a strategic decision, not a technical one: each option has a different impact on visibility, content protection, and subscription conversion.

- Anticipate AI: allowing Google to read your content also means opening the door to ChatGPT or Perplexity; blocking Google risks a drop in SEO.

- Secure markup, and calibrate text before the paywall to prevent content leakage and find the balance between SEO (≥ 800 characters) and conversion.

- Adopt a hybrid blocking model: adjust the level of openness according to the value, lifespan, and type of content, rather than a single rule.

- Don't confuse protection with growth: blocking AI protects your content, but only a strong brand and a clear offering will create subscribers.

The post Paywalls & SEO in the AI era appeared first on Audiencers.

David Carzon (formerly of Arte, Télérama, Libération, and Binge Audio) is a journalist, consultant, and author. This article was originally published in his newsletter Hupster, and provides a follow up from Audiencers first article on Le Monde in English, published six months after its launch in 2022.

Increasingly more publishers are launching international editions, all looking to capitalize on the potential of AI translation, particularly into English. While these announcements sound promising, I have my doubts about their viability for private media outlets.

To see if there’s real opportunity here, I sat down with Arnaud Aubron, Director of Diversification at Le Monde. Having led Le Monde in English project for four years now, he shares why they launched, the target audience, the work required for this project, and how it’s on track to become profitable in 2026.

Does that make it a replicable blueprint? It’s not that simple. I’ll let him explain…

> Le Monde in English expects to be profitable in 2026, four years after launch rather than the five projected. It got there by growing to 15,000 subscribers and dropping the translation agency it started with

> The workflow is AI first, humans after. Every article is machine-translated, then proofread, then edited by a native English-speaking journalist, with eight of them across Paris and Los Angeles for 24-hour coverage

> AI didn't cost jobs here, it created ten permanent journalism posts. Arnaud Aubron is explicit that without it the project would have been too ambitious to attempt at all

> The product is deliberately a second read. Prices are kept low because nobody is dropping their primary title for this, and students and retirees are overrepresented among subscribers

> The warning for anyone copying it: this works because Le Monde already had brand recognition in the US and UK, and because French to English opens the American market. As Aubron puts it, "the news market remains stubbornly national"

What needs did you identify that led to the launch of Le Monde in English?

The post Profitable by 2026: the business logic behind Le Monde in English appeared first on Audiencers.

Morten Stald is a Senior Partner at Subscrybe, the leading Scandinavian consultancy specializing in subscription. In this article, Morten shares why he thinks you need to update your subs management system, what you should do before moving forward and what you need to consider when selecting the replacement solution.

We live in a digital age. Regardless of the type of subscription you offer, you likely have numerous digital interactions with your customers. This places high demands on the system infrastructure that supports your subscription business. At the heart of that infrastructure is a ‘subscription management system‘ which essentially handles the management of your subscribers, pricing, and subscription plans, ensuring that customers are charged the right amount – every time.

But, as a media professional, you already know that. However, do you know what is expected of your media business in the near future?

We keep a close eye on what subscribers like and don’t like. And we know for a fact that subscribers are harder to satisfy than transactional customers. After all, you need a good reason for charging them every month – and you need to deliver more and more value each month, in order for them to stay.

This is why we are currently helping Nordic media companies choose the best new subscription management system for their needs. And, if you didn’t already do it, we strongly encourage you to consider it as well. Because a new media reality is here. One, where your digital capabilities must be second-to-none, if you want to retain your readers over time. As print circulation slowly declines, media companies must prepare for delivering the most value in a new digital arena. And that is hard to do, without strategic IT investments.

The post Why it’s time to update your subscription management system to accommodate consumer trends appeared first on Audiencers.

Jason Langsner is Group Product Manager, Data + AI, at The Washington Post. In this article he shares how they built The Post’s in-house generative AI chatbot tool that uses RAG to answer user questions.

> The Washington Post built Ask The Post AI, a chatbot answering reader questions using retrieval augmented generation over its own reporting from the past eight years, with links to every source article

> It started as one desk rather than a platform. Climate Answers launched with the Climate Desk, drawing only on Post climate coverage, before the approach was scaled site-wide using Meta's open-source Llama models

> The hard problem isn't answering. It is teaching the model to say when The Post hasn't covered something well enough to give a solid answer, which is the work Jason Langsner singles out

> It came out of a five-day AI hackathon in 2022, run by engineering and product alongside the newsroom. That became a permanent R&D lab, and The Post now has a Chief AI Officer

> Internal first, external second. The lab builds tooling for Post staff as its own customers before features reach subscribers, so the newsroom stress-tests them before readers do

The Washington Post may be best known for our world-class and award-winning reporting, but we also pride ourselves on being industry leaders in technology and innovation. In addition to producing journalism, we also develop products and solutions in house to improve our users’ experience, including through our expanding AI Pod Research & Development (R&D) lab and newly launched WP Incubator.

The post The Washington Post: how and why we built “Ask The Post AI” appeared first on Audiencers.

This article summarizes the session given by Max Leroy, Audience & Product Strategist, formerly of The New York Times, CNN, and POLITICO, at The Audiencers’ Festival in Paris on September 16th, 2025.

This session in 5 bullet points:

• The real danger to the media? Not AI or Google, but our disconnect from readers’ needs.
• Plummeting trust, dwindling audiences, refusal to pay: in any other industry, we would be talking about the collapse of product-market fit.
• Urgent action is needed: we must stop pointless obsessions and return to users’ needs.
• This series of articles presents five toxic obsessions and offers five useful obsessions for the sustainability of the media: short circuits, communities, respect for audience needs, experimentation, and advertising/UX balance.
• This second chapter covers obsession number two: refusing to interact with audiences

What if the greatest threat to the media were neither AI, nor the “big bad wolf” Google, nor the decline in social traffic, not even the crisis in publishers’ business model? What if the real problem were us? Us and our persistent inability to take an interest in those we are supposed to serve: our readers.

The post The media industry’s 5 toxic obsessions: no.2 refusing to interact with audiences appeared first on Audiencers.

Ali Mahmood is a Digital News Strategist, helping publishers navigate the evolving digital landscape to build sustainable, audience-first revenue strategies—even in the most challenging environments.

This article was written off the back of a conversation from The Audiencers' Whatsapp group about how to prevent password sharing and account misuse within organisations. 

Many news and media publishers reach a point where subscription growth stalls even when all the basics seem right – a strong product, smooth onboarding, effective newsletters funneling readers in. Often an unseen culprit behind this plateau is revenue leakage: people accessing paid content without paying (or without paying correctly). This leakage tends to hide in the data – for example, odd spikes in “active users per account,” lots of free-trial users who never convert to paid, or unexplained churn. If left unaddressed, it quietly undermines growth, cutting into revenue and even skewing audience metrics. It can also create fairness issues that erode trust between your newsroom and paying subscribers.

Why revenue leakage happens (and why it hurts)

The post How to stop subscription revenue leaks: technical and strategic solutions appeared first on Audiencers.

Over the past few months, the team at Hearst Newspapers has been experimenting with AI tools to create “Connecticut Today”, CT Insider’s weekday audio news briefing. 

At first the team thought they could summarize a few stories, top with a quick introduction and finish with a closing sentence or two. Those initial scripts were fine but not really something you’d want to listen to daily to get caught up on the news. 

The post Launching a weekday audio news briefing with AI at CT Insider appeared first on Audiencers.

I'm Ali Mahmood, and I've spent 20 years watching the media industry make the same costly mistake: confusing technological activity with strategic impact. As someone who's led digital transformations at news organizations across four continents, I've seen too many smart people chase the latest platform update while their audiences quietly slip away. I'm here to help separate signal from noise!

Media has become a technology business. This isn’t rhetoric, it’s operational reality. Whether you’re running a century-old newspaper or a digital startup, your survival depends on how effectively you deploy technology to reach audiences and generate revenue. As newsrooms face mounting pressure from creator platforms and AI systems, the decision to build custom technology or buy existing solutions has moved from the margins to the center of strategic planning.

Understanding software development lifecycles and conducting rigorous cost analysis aren’t optional skills anymore. They’re survival tools. The industry is littered with expensive failures—custom CMS projects that consumed millions and delivered nothing, ad tech adventures that ended in obsolescence. But there are also remarkable successes where proprietary technology created lasting competitive advantages.

The essential frameworks

The Software Development Life Cycle( long read here) sounds abstract but it’s simply how software gets built in the real world. Understanding its phases helps you spot where projects typically fail and why timelines explode:

• Planning and Requirements: This is where you define what you’re actually building and why. In newsrooms, this phase often gets rushed because everyone thinks they already know what they need. They rarely do. Requirements shift, stakeholders multiply, and scope creeps before a single line of code is written.
• Design: Architecture decisions made here determine whether your system scales or collapses under load. User interface choices affect adoption rates. Database structures impact performance for years. These aren’t just technical decisions—they shape how journalism gets done.
• Development: Where code gets written. Usually the most visible phase and paradoxically the most predictable. Good developers can estimate coding time reasonably well. It’s everything around the coding that destroys schedules.
• Testing: Reveals the gap between intention and reality. Every newsroom that’s built custom tools has stories about features that worked perfectly in development but failed spectacularly when journalists actually used them. Proper testing doubles timelines but prevents disasters.
• Deployment: Moving from controlled environments to production systems where real journalism happens or where distribution happens surfaces integration issues, performance problems, and workflow conflicts that nobody anticipated.
• Maintenance: The phase that never ends. Software isn’t a product you finish it’s a service you provide. Bugs need fixing, features need updating, security patches need applying. The true cost of custom software lives here, accumulating month after month, year after year.

Total Cost of Ownership captures this full lifecycle burden. It’s not just what you pay to build—it’s what you pay to run, maintain, upgrade, and eventually replace. Operational expenses typically dwarf upfront investments. I’ve seen newsrooms spend five times their initial development budget just keeping custom systems running over three years.

The strategic trade-offs

Arguments for buying:

• Speed and predictability—vendors have already absorbed development costs and technical risk
• Your team focuses on journalism instead of debugging code
• Access to ongoing updates and support from specialists
• Proven systems that work from day one

But buying means accepting constraints:

• Limited customization options
• Your data lives in someone else’s system….This is what the first party data is discussion is partially about.
• Dependence on vendor roadmaps and pricing decisions.
• Generic workflows that might not match your newsroom’s needs

Arguments for building:

• Perfect alignment with your specific workflows
• Complete control over features and roadmap
• Ownership of code and data or open source. Open source means somebody did some serious work that you won’t need to pay for.
• Potential competitive advantage through unique capabilities

But building demands sustained commitment:

• Technical talent that understands both code and journalism. Bridge roles become important.
• Deep pockets for development and maintenance
• Tolerance for risk and potential failure
• Years-long commitment to ongoing development

The new complexity

AI adds urgency to these decisions. Should you develop proprietary AI tools or integrate existing services? The same calculus applies: What’s core to your value proposition? What resources can you commit? Where will proprietary capability create lasting advantage?

The creator economy raises the stakes. Individual creators now have access to sophisticated tools from major platforms. News organizations must match or exceed these capabilities while maintaining journalistic standards.

A strategic framework

The build versus buy decision isn’t binary, it’s a spectrum of options. As Aakash Gupta argues in his product leader’s guide, these decisions should align with your strategic priorities and organizational capabilities. The International News Media Association’s recent analysis similarly highlights this “gray area” between pure building and buying.

Build when:

• The technology is central to your unique value proposition
• You have resources to sustain development over years, not months
• Off-the-shelf solutions fundamentally don’t meet your needs
• The competitive advantage justifies the risk

Buy when:

• You need proven capability quickly
• The function, while necessary, isn’t your differentiator
• Your resources are better spent on content and audience development
• Vendor solutions meet 80% or more of your requirements

The hybrid path often makes most sense: buy core platforms but customize through APIs and open source integrations. This balances control with efficiency. Modern vendors increasingly offer this flexibility, understanding that no single solution fits every newsroom.

Without technical leadership at the strategic level, even established media organizations risk irrelevance. The future belongs to those who treat technology decisions as editorial decisions because in digital media, they’re the same thing.

The post Build or buy: The technology decision that defines modern newsrooms appeared first on Audiencers.