I spent the first two days of October at Circa in Amsterdam for ContentCon Europe 2026. The event theme was ‘From here to agentic: your roadmap for an adaptive brand’. It was a great event, with some exciting announcements from Contentsack about upgrades to their digital experience platform. However, the most useful material came from the customer stages. Reckitt, Philips, Mondelēz and GetYourGuide all spoke openly about what they have built, where the hard work sat and what changed as a result. iO and a partner panel explored the related challenge of how brands are found and represented by AI agents.
Taken together, the sessions showed that meaningful progress comes from treating structured content, operating models, machine readability and human judgement as connected concerns. I came away with four priorities for digital leaders to act on now:
Structured content is the new business infrastructure
Your digital content now has two audiences: people and agents
Simplification creates speed at scale
AI creates value where it strengthens human judgement
The common thread was to structure the content, simplify the system, make it legible to machines and keep people accountable for what reaches the customer.
1. Structured content is becoming business infrastructure
The clearest message from ContentCon was that content structure now affects far more than publishing efficiency. It shapes whether information can be reused, localised, personalised, governed and understood by AI. iO’s Erik Post offered a useful test for any content model: can a piece of content stand on its own without the page it sits on? If it can, it has a chance of working across a website, an app, a retailer feed or an AI-generated answer. If it can’t, the the content remains tied to the layout and channel it was created for, limiting its usefulness to an organisation.
Philips, meanwhile, showed what happens when flexibility grows without enough structure. Its marketers had a large component library and considerable freedom to create content in whatever structure they wished. However, this eventually produced 1.9 million pages across 90 locales, including around half a million unstructured pages. The solution was a much smaller set of 35 purposeful templates, supported by consistent content models and taxonomy.
That foundation also changed the economics of content migration. Philips combined page data, analytics, SEO value, business input and legal requirements, then used AI to recommend which pages to keep, merge or remove. After humans assessed the output, around 900,000 redundant pages were removed. The result is a smaller, more manageable estate that can support personalisation, syndication and automated workflows because the underlying content is coherent.
Reckitt reached a similar conclusion at a different scale. It consolidated hundreds of sites and multiple content management systems onto a shared platform, underpinned by a design system, brand tokens and reusable content. The platform gives central teams control over security, accessibility, quality and performance, while brand teams retain ownership of content, look and feel and the customer journey. Clear structure created room for both governance and speed.
Learn how Inviqa helped build Reckitt’s global design system that supported this consolidation.
2. Digital content now has two audiences: people and agents
Brands still need to persuade people, but increasingly AI agents decide whether people encounter a brand at all. iO described the risk as content drift. In research covering 37 brand questions across ChatGPT, Gemini and Perplexity, most answers were correct, but 62% of the incorrect answers were out of date and 11% were fabricated. Expired offers and obsolete company information no longer confuse one visitor at a time. They can be repeated across many AI-mediated conversations before the brand realises there is a problem.
The partner panel with Huge, MSQ and Cloudinary made the practical implications clearer. Agents struggle with sites full of code focused on how content on the page looks rather than what that content means and media that carries little descriptive metadata. A carefully art-directed photograph or video has limited value in an agentic journey if the machine cannot understand what it contains. Clean markup, structured data and enriched asset metadata are now crucial both for both the on-page experience as well as to improve discoverability.
This changes the role of the website too. Discovery may happen elsewhere, with an agent forming the shortlist before a customer arrives. The people who then do visit are likely already closer to making a decision, raising the stakes of landing pages, product pages and the credibility of the experience. The machine-readable layer and the human-facing experience need to be designed as one system.
GetYourGuide added that its research found people use AI for recommendations while still wanting evidence that a real person has visited a place and can speak from experience. That authentic editorial content was also being cited by Gemini and Google AI Overviews. Human expertise gives people confidence and gives AI systems a source worth referencing.
AI visibility is becoming a measurable content-quality problem. Organisations can start by asking the major models about their own brands, products and offers, then tracing inaccurate answers back to stale, unstructured or poorly described source content.
3. Simplification creates speed at scale
The strongest transformation stories at ContentCon were about reducing unnecessary choice. Reckitt once had around 400 websites across as many as 12 content management systems. Following the consolidation of its platforms, migration speed increased from one site a month to 25, while the cost of site builds and post-launch publishing halved. Shared templates and components allowed accessibility, performance, SEO and privacy standards to be applied consistently across the estate.
The operating model mattered just as much as the platform. Reckitt standardised workflows and established a dedicated delivery team with clear ownership. Rollout cycles dropped from months to weeks, and some brand approvals now return within 30 minutes. The lesson is relevant to any organisation seeking more speed from its agencies or internal teams: continuity, visibility and decisions close to the work create a stronger delivery system.
Philips simplified in a similar way. Reducing a sprawling component landscape to 35 core templates made journeys more consistent and changes easier to propagate. One update can now reach hundreds of pages that previously required repeated manual editing. Image handling was simplified too, with one focal point used to generate crops for different screens rather than producing nine versions by hand.
GetYourGuide identified its constraint by mapping the editorial workflow rather than assuming authors needed to write faster. Editors were losing time translating marked-up Google Docs into CMS components. Its internal AI tool now reads the document and publishes it to Contentstack using the correct structure. Output rose from around 30 articles a month to 500, while writers remained in familiar tools and editors retained control of quality.
Across all three examples, speed came from removing friction around the work. The technology helped, but only after the organisations had narrowed the choices, clarified the patterns and identified the real bottleneck.
4. AI creates value when it strengthens human judgement
None of these stories present AI as a substitute for ownership. iO’s phrase, ‘agents advise, editors decide’, captured the prevailing pattern. Its Content Raider concept watches external signals, identifies content that may have drifted and asks an editor to approve the change. The automation finds and prepares the work, while a person remains accountable for what is published.
Philips used the same principle during migration. AI recommended whether content should be kept, merged or removed, with stakeholders confirming the decision. GetYourGuide used AI to take friction out of publishing and localisation rather than generate a volume of anonymous copy. The combination of machine translation followed by AI post-editing outperformed either approach on its own, while the source material continued to rely on genuine human experience.
Mondelēz showed how this thinking applies to personalisation. Its ‘empathy at scale’ strategy connects paid media, web, CRM and customer service data so experiences can respond to the individual rather than the channel. The reported commercial gains were substantial, including 70% incremental revenue from first-party audiences in paid media, a 44% lift in website sign-ups and higher engagement across web and email.
However, Mondēz’s Kim Pham was clear about the organisational reality. The programme went over time, over budget and out of scope, and the hardest work involved sponsorship, process and winning the confidence of brand managers. Her advice is to begin with a single use case and a single channel, prove value quickly, build in contingency and stop what’s not working rather than attempting to transform every journey at once.
The repeatable pattern is to use AI for detection, recommendation, assembly and scale, while keeping human expertise, approval and accountability in the loop.
What I encourage organisations to adopt
These stories suggest a practical roadmap for organisations that want to move from AI experimentation to dependable digital capability.
Check how AI describes your brand. Ask the major models about your products, services, prices and offers. Turn incorrect or outdated answers into a prioritised content-quality backlog.
Model content around meaning. Give every content type a clear purpose and owner, then use the stand-alone test to check whether it can travel beyond its original page or channel.
Simplify before you automate. Reduce unnecessary templates, components and workflow choices. Automation compounds the quality of the system it is introduced into.
Use AI to remove the actual bottleneck. Map the work first. Apply AI where it can reduce repetitive handling, surface decisions or accelerate localisation while preserving expertise and accountability.
Create clear ownership across technology, content and brand. Shared platforms work when central standards and local decision rights are explicit, supported by stable teams and transparent delivery.
Start personalisation with one valuable use case. Choose a channel where the organisation has usable data and a measurable outcome. Prove the value, learn from the result and expand deliberately.
A final thought
ContentCon Europe reinforced that the next stage of digital experience will depend less on adding isolated AI features and more on improving the foundations around them. Structured content, legible data, simpler platforms and accountable operating models give organisations the ability to adapt without losing control or trust.
If you’re exploring AI visibility, structured content, large-scale platform consolidation or more adaptive customer experiences, get in touch. Inviqa can help turn these ideas into a practical plan and working capability. Get in touch today.
