Beyond the CMS - A Content Management Podcast

What happens when website search evolves from a keyword box into an intelligent interface?

In Episode 50 of Beyond the CMS, Chris Bryce speaks with Olivier Dobberkau, CEO and founder of dkd Internet Service GmbH and president of the TYPO3 Association, about how enterprise search is moving from keyword matching toward semantic search, retrieval-augmented generation (RAG), conversational experiences and agent-readable interfaces.

Olivier explains why AI search begins with content quality, metadata and retrieval. They discuss Apache Solr for TYPO3, embeddings, document chunking, grounding, hallucination risks, search feedback and the difference between a generic chatbot and specialized search capabilities routed by user intent.

In this episode:

• Why search is becoming an interface rather than only a search box
• How keyword search differs from semantic search and RAG
• How Apache Solr supports TYPO3 search
• Why poor content creates poor AI-search results
• How chunking can remove important context
• Why specialized search capabilities can outperform generic chatbots
• How real questions, zero-result searches and user feedback improve retrieval
• Why organizations should repair content before adding generated answers

Chapters

0:00 Welcome to Beyond the CMS Episode 50
2:06 Olivier Dobberkau, dkd and the TYPO3 ecosystem
5:32 Apache Solr search for TYPO3
7:23 Why search is becoming the interface
11:21 From keywords to conversational and agentic search
13:47 Content quality, grounding and search quality
19:46 Why chatbots are not always the right experience
22:37 Specialized search capabilities and intent routing
24:02 Question-based website search in practice
26:19 Repairing content before adding RAG
28:33 Retrieval, metadata, feedback and supervision
31:25 RAG explained: indexing, embeddings and retrieval
34:16 Chunking, context and grounded answers
36:39 Open source, access and the TYPO3 community

About Olivier Dobberkau

Olivier Dobberkau is CEO and founder of dkd Internet Service GmbH and president of the TYPO3 Association. His work spans open-source CMS development, TYPO3 search, Apache Solr and the transition from keyword search toward intelligent content discovery.

Connect with Olivier on LinkedIn
Learn about dkd's TYPO3 work
Learn about Dotfusion

Most accessibility tools tell you what is broken. Limena helps teams understand which accessibility barriers are disrupting valuable customer journeys, what those failures may be costing the business, and who owns the fix.

Ryan Short, founder of Limena, joins Chris Bryce to explain how accessibility intelligence turns a long list of WCAG failures into business priorities. Limena connects accessibility barriers to customer journeys, conversion points, organizational ownership, and continuous verification.

They discuss:

• Why accessibility work keeps losing to the product backlog
• How to connect accessibility barriers to customer journeys and revenue
• How Limena prioritizes issues by business impact
• Why closing a development ticket does not prove the barrier is gone
• Why widgets and point-in-time audits are not enough
• Where accessibility overlaps with SEO, AEO, and content quality
• How Limena fits into Git-based and composable delivery workflows

Watch the episode: YouTube

Learn more about Limena: limena.app
Connect with Ryan: LinkedIn
Learn more about Dotfusion: dotfusion.com

McGill University runs a thousand sites on its central platform, 500 more custom builds, and tracks 6,000 to 8,000 content edits a week. Joyce Peralta, Manager of Digital Communications at McGill, joins Chris Bryce to talk about the content model project addressing all of it.

The number that anchors this conversation: 61 different terms used across the university for the exact same program component. Joyce's central point is that AI does not cause content inconsistency, it amplifies inconsistency that already exists. The inaccuracy people see in AI answers comes from information presented differently across websites, not from information that is wrong.

McGill looked at building a separate content repository just for AI and turned it down, because a second repository means a second governance and synchronization problem. Joyce points to Dries Buytaert's markdown experiment, which found AI bots still hit the website itself the majority of the time.

The content model work is federated across about 50 subject matter experts from departments across the university. Rollout runs concept by concept, with a timeline Joyce estimates at five years or more. Her real point on success: this cannot be a Communications project or an IT project alone, it has to be both working as one team, with alignment across the whole institution.

Also in this one: the student usability panel that tracks how AI use among students changes year over year, why Drupal and the Drupal EDU Slack group are a real source of leverage, and what a taxonomy problem looks like at university scale.

Chapters

0:00 Welcome to Episode 48
 1:05 Meet Joyce Peralta, McGill University
 1:38 The velocity of change since Beyond the CMS launched
 2:35 Inside McGill's web services team: comms and IT as one team
 4:12 Who owns the university's content model
 5:56 The vision behind McGill's content transformation
 9:34 Taxonomy at scale: a thousand sites, 500 custom builds, 6,000 to 8,000 edits a week
 14:33 Federated with 50 subject matter experts: the 61-term problem
 16:19 AI amplifies inconsistency, it does not cause it
 17:19 Why McGill rejected a separate AI-only content repository
 19:10 How McGill actually uses AI in the content model project
 21:35 Concept by concept: a rollout Joyce estimates at five years or more
 24:02 Open source, Drupal, and the Drupal EDU Slack group
 26:49 Three feedback groups, including the student usability panel
 29:50 The human side of technology as AI reshapes expectations
 31:32 Lessons for higher ed teams just starting this work
 33:38 Where to find higher ed thought leadership
 34:52 Wrap-up

Click here to watch a video of this episode.

Links
Joyce Peralta on LinkedIn
Dotfusion headless CMS services
Chris Bryce on LinkedIn

Click here to view the episode transcript.

Enterprise content supply chains run 15 to 19 steps. Joel Varty, CTO of Agility CMS, thinks AI can get that down to six, so people are the only bottlenecks left. Chris Bryce and Joel dig into the MCP server that connects Claude directly to your CMS and kills the copy-paste layer of content operations.

Joel has spent 21 years building Agility, the Canadian headless CMS behind Cineplex, Scotiabank country sites, school boards, and higher education. He explains why Agility deliberately did not build an AI agent inside the CMS, and instead meets teams inside the AI they already use: Claude, Copilot, or Gemini.

The MCP server now runs the full loop: content models for developers, a content pipeline that pushes up to 250 items per batch, and bulk publish, unpublish, and delete. Plus metadata vs JSON-LD schema in plain language, AEO, blogging from a phone with a Claude project, and content ops through a CLI for regulated industries that need sign-off and audit logs.

Click here to watch a video of this episode.

Links: Agility CMS: https://agilitycms.com Connect your AI agent to Agility: https://mcp.agilitycms.com Dotfusion (Agility Partner): https://dotfusion.com/services/agility-cms-development-agency Chris on LinkedIn: https://www.linkedin.com/in/cbryce Joel Varty on LinkedIn: https://www.linkedin.com/in/joelvarty/

Creators & Guests

Click here to view the episode transcript.

Most enterprise sites are not ready to be found by AI search, and the fix starts before you ever migrate. Chris Bryce talks with Joel Goodman, VP of Growth Strategy at Squiz, about content intelligence, query fanout, and what makes a composable DXP AI-ready.

Joel spent 13 years running an agency in higher education before joining Squiz, a 26-year-old Digital Experience Platform built in Australia with a growing North American presence. We get into what makes Squiz different from a traditional headless CMS, why accessibility is now the same problem as AI discoverability, and how Squiz Content Intelligence audits your content the way an AI search engine reads it.

The sharpest idea: query fanout. Ask one question in ChatGPT and it breaks into 40, 50, sometimes 60 sub-queries your content has to answer, queries you never see.

Click here to watch a video of this episode.

Links: Squiz: https://www.squiz.net
Dotfusion (Squiz Partner): https://www.dotfusion.com
Chris on LinkedIn: https://www.linkedin.com/in/cbryce
Joel Goodman on LinkedIn: [ADD URL]

Creators & Guests

Click here to view the episode transcript.

Legacy CMS platforms lock your front end and back end together, so every new channel becomes its own project. In this live AMA, Chris Bryce sits down with Luke Wertz and Nick Switzer of Contentful to talk about what it actually takes to move off WordPress, Drupal, Sitecore, or AEM and into a modern, composable content stack.

Luke manages Contentful's commercial mid-market solution engineering team in North America. Nick manages the enterprise side from Denver. Between them they have close to four decades in the CMS space, and they walk through Contentful's founding story in Berlin, where updating a single word in a native mobile app used to mean shipping a whole new version of the app.

We get into the honest difference between legacy, monolithic, and composable platforms, what a migration actually looks like in 2026, why content modeling is the foundation that makes everything downstream faster, why Contentful stays rigid on structured JSON delivery while building flexible tools for editors, and how Contentful's AI Actions and API first management layer let agents like Claude create, scale, and govern content directly inside the platform.

• (00:00) - Welcome to Episode 45: the "Car Guys" origin story

Click here to watch a video of this episode.

Links: Contentful: https://www.contentful.com Dotfusion (Contentful Partner): https://dotfusion.com/services/contentful-development-agency Chris on LinkedIn: https://www.linkedin.com/in/cbryce Luke Wertz on LinkedIn: https://www.linkedin.com/in/lukewertz/ Nick Switzer on LinkedIn: https://www.linkedin.com/in/nickswit/

Creators & Guests

Click here to view the episode transcript.

Derk Majoewskij is Senior Product Strategist at Storyblok, where he leads product discovery and roadmap prioritisation across the full product cycle. In four-plus years at the company he has worked across product management and product marketing, shipping features end-to-end from problem discovery through engineering delivery to go-to-market. He was the primary voice behind FlowMotion, Storyblok's native automation and orchestration layer built on n8n, and has been closely involved in Storyblok's AI direction including the MCP server and the Strata semantic vector layer.

The conversation starts where most CMS conversations don't: not at the platform decision, but at what happens to content after the site goes live. Derk and Chris dig into the gap between a CMS that holds content and one that actively operates it, and why that distinction is becoming the difference between brands that show up in AI-driven search and those that don't. Derk frames it directly: content used to be static and humans made every change. The shift underway moves the human into a loop where agents, automation, and semantic indexing do the operational work.

Chris walks through Dotfusion's own 140-node agentic content workflow, from AEO-informed topic selection through research, writing, governance, schema refactoring, and social distribution, all orchestrated through n8n and Storyblok. Together they cover the real upfront cost of building AI content infrastructure properly, why automation native to the CMS handles enterprise IT requirements that external tools can't, and how voice-to-content workflows are already running in production for franchise and retail clients.

Key topics include:

• FlowMotion: Storyblok's n8n-based automation layer, built natively inside the CMS with custom Storyblok triggers and actions already in place. One publish event can trigger translation, republishing, and a Slack notification with no external setup required — and no IT overhead for enterprise teams.
• Strata: Storyblok's built-in vector database, in active development, that makes an entire content history semantically readable by AI tools rather than just keyword-searchable. Essential for any brand managing GEO visibility across years of published content.
• MCP Server: Storyblok's Model Context Protocol integration that connects AI models directly to a Storyblok space, enabling natural language content operations, story search, and content updates without leaving the platform.
• The 25x rule: Building AI content infrastructure properly takes roughly 25 times longer than writing a single post. The argument is that teams investing now build a permanent operational advantage that compounds while others are still starting.
• Living vs. static content: Derk's framework for the shift from a CMS where humans manage every update to one where content is automatically updated, optimised for AI search, and made semantically available to agents while the human remains in the loop.
• Voice-to-content in production: How franchise and retail brands are using a phone, a store ambassador, and a FlowMotion governance model to turn raw audio captured on the floor into polished, on-brand content without stripping the authentic voice out of it.

The infrastructure decision you make now about your CMS determines whether your team is spending its time on strategy or on operations, and whether your content is visible to AI tools or invisible to them.

🎧 Listen now and join the conversation!

Preston So is the Chief Commercial Officer at React Bricks and the founder of Decoupled Days, a nonprofit conference on headless CMS, AI, and content operations now in its ninth year (August 6-7 in Montreal). Over 22 years in the CMS industry, spanning Oracle, Acquia, Drupal, platforms written in Java and PHP, and now React-native tooling, Preston has watched every architectural wave move through enterprise content teams. His perspective on where things are actually breaking right now is one of the sharpest in the market.

This conversation lands on a problem most enterprise teams are living but not quite naming: the collaboration breakdown between designers, developers, and content editors. Both monolithic DXPs (priced at six or seven figures, bundled with tools most organizations never touch) and composable headless platforms (fragmented, developer-heavy, expensive to integrate) have failed to maintain what Preston calls the "through line" — the shared context that connects design systems, content, and code. Now that AI agents need to operate across all three simultaneously, the absence of that through line is no longer just an inconvenience. It is an architectural gap.

React Bricks answers this with what they call a design system native CMS: built on modern React frameworks like Next.js and Astro, with visual in-page editing that keeps marketers within brand guardrails, and AI features shipping this month that generate complete landing pages from a short prompt — not just individual text fields — while keeping every component tied to design tokens and brand colors. Preston demos this live in the episode, including a real-time page build and a hiccup that makes it even more honest.

Key topics include:

The composable overcorrection: Headless platforms solved vendor lock-in but created a new problem: fragmented architecture and a collaboration layer no one fully owns. Preston argues the market is now swinging back toward something with more structural cohesion.

Design system native CMS: Treating the design system as a first-class citizen in the CMS, not an afterthought. This matters for marketers who need brand consistency and for AI agents that need a stable, structured context to operate within.

Vibe coding has a ceiling: AI-assisted website generation works well for a static site you are willing to rebuild from scratch. A colleague of Preston's recently shared data with a clear inflection point: once a company reaches $1M ARR, the clean-sheet approach breaks down. Stability, governance, and design system fidelity become non-negotiable.

AI page generation vs. AI field generation: Text field automation, alt text, metadata — these are table stakes now. Every CMS does them. The harder frontier is generating complete pages within a design system from a short natural language prompt. React Bricks is shipping this now.

The through line problem: Both monolithic and headless CMS architectures severed the connection between designers, developers, and content editors in different ways. With AI agents now operating at every junction, that connection has to be rebuilt deliberately. The CMS is the layer where it has to happen.

Decoupled Days 2025: The ninth annual nonprofit conference on headless CMS, AI, and content operations runs August 6-7 in Montreal. Dotfusion is a sponsor and Chris Bryce is speaking. Tickets are $35 CAD (early bird, available while sessions are still being announced). decoupledays.com

Preston's central argument: as AI takes over more of the production layer, the collaboration infrastructure — who can edit what, within what constraints, across which personas — becomes the most important architectural decision a team can make. That is what the CMS has to solve for now.

🎧 Listen now and join the conversation!

In Episode 42 of Beyond the CMS, host Chris Bryce, founder of Dotfusion, is joined by Joel Varty, CTO at Agility CMS. Joel recently celebrated his 21st anniversary at Agility — a bootstrapped, Canadian headless CMS that has never taken a dollar of venture capital and still counts customers from nearly its founding days. That long arc gives him a uniquely grounded perspective on where the CMS industry has been and where it is actually going, as opposed to where investors want it to go.

The conversation opens on a question that sounds simple but isn't: what is headless, really? From there, Joel walks through why enterprises are finally abandoning monolithic DXP platforms en masse, why the composable model the MACH Alliance promotes is different from those platforms trying to rebundle the same features under a new name, and why the best AI workflows are being built around headless systems right now. He also shares a live demonstration of publishing a personal blog post directly from Claude on his phone, straight into Agility CMS, using an MCP server connection and a pre-built skill — no copy-paste required.

Key topics include:

• Why bootstrapped beats venture-funded for CMS longevity: Agility has never taken outside capital, which means product decisions are driven by customer needs rather than growth-at-all-costs mandates. Joel argues this is directly why their average customer lifetime exceeds 15 years.
• Headless as the AI-ready architecture: Adding AI onto an API-first, headless system is dramatically easier than retrofitting it onto a monolithic DXP. Every headless CMS has been able to take advantage of MCP servers and AI agents specifically because content and code are already separated.
• MCP: the layer that teaches AI how to use your tools: The Model Context Protocol is not just a wrapper on an API. It contains human-readable instructions that explain what every field type actually means — so AI agents can act on your CMS without a developer interpreting the API for them.
• Skills as the next standard for AI agents: Joel predicts that Claude-style "skills" — pre-configured prompt contexts scoped to a specific task like "create blog post" — will become a universal standard. His team has already built org-wide skills that pull simultaneously from Jira, Figma, and Agility MCP servers.
• Content Operations, redefined: The content operations pipeline has always been idea to draft to CMS to publish. What's changed is that AEO (Answer Engine Optimization) and GEO have replaced the SEO ranking game as the reason content quality matters. AI models cite based on truth and structure, not keyword density.
• Why DXPs are creating the problem headless was built to solve: Several major DXP vendors are now acquiring personalization, search, and analytics tools to capture more revenue — which is exactly the monolithic bloat that composable architecture was designed to escape. Joel says Agility is staying focused: content, done exceptionally well.
• Data sovereignty as an enterprise buying signal: Canadian government websites, Scotiabank's Latin American properties, and Cineplex are all on Agility in part because of data residency control. With CFOs and legal teams now actively asking where AI tools process their data, this is becoming a board-level conversation, not just an IT checkbox.
• The CMS as a diamond: Agility's product philosophy is subtraction. Their director of product frames the CMS as a diamond — and the job is to keep chipping edges off. Every recent update removes steps rather than adding features. Joel's example: a locales update that reduced the number of clicks for multi-locale teams, which sounds minor until you manage 30 locales like Epson does.
• Stop making marketers fill out fields: The long-form piece is the only thing a marketer should have to write. AI should handle field population, metadata, and content model mapping automatically. When publishing is that frictionless, teams actually produce more content — Joel went from occasional to blogging two or three times a week.

Joel's bottom line is that the CMS is not where AI should live. AI should live in the agents your teams already use — Claude, ChatGPT, Gemini — and those agents should connect to your CMS through MCP. The CMS is the foundation. The cleaner and more structured that foundation is, the more powerful every AI workflow built on top of it becomes.

🎧 Listen now and join the conversation!

In Episode 41 of Beyond the CMS, host Chris Bryce, Partner at Dotfusion, is joined by Mike Vertal, CEO of Crafter CMS. Crafter is an enterprise-grade headless content platform running major digital properties for brands like eBay, Papa John's, and Marriott. Before building Crafter, Mike co-founded a digital agency similar to Dotfusion, grew it to serve large enterprises, and sold it to Capgemini — giving him a rare, dual perspective on both the agency side and the CMS product side of the industry.

The conversation starts with what makes Crafter architecturally different from every other headless CMS: it stores all content in Git instead of a database. From there, Mike walks through exactly why that decision matters now that AI has entered the picture, and how it shapes the way authors, developers, and site visitors all experience a digital property. He also introduces Crafter Q, a brand-new standalone product that puts a custom, brand-controlled AI conversational agent on any website, regardless of the CMS powering it.

The discussion goes well beyond product features and into the broader strategic questions every digital team is wrestling with right now: How do you prepare your CMS for AI? Where does your content live versus where should it live? And as browsing gives way to searching, and searching gives way to conversing, what does that mean for the websites you're building today?

Key topics include:
• Git as a Content Repository: Why building around Git instead of a database removes environment-syncing headaches, improves security and scalability, and turns out to be naturally AI-friendly.
• Headless vs. Decoupled: The important distinction between decoupling the front-end presentation from the back-end versus decoupling the authoring system from the delivery system entirely.
• Headless Plus: The extra capabilities Crafter layers on top of a basic content API: embedded search, server-side scripting, and a full WYSIWYG authoring experience.
• AI for Content Authors: How MCP (Model Context Protocol) integration and a native AI chatbot inside the authoring interface help teams draft content, populate metadata, and stay current with SEO and GEO best practices.
• AI for Developers: Using AI coding tools like Cursor, enhanced with CMS-specific AI skills, to go from HTML template to a fully CMS-enabled site in minutes, and collapsing migration timelines from months to days.
• AI for Site Visitors: Why every website will eventually need a custom conversational agent trained only on your content, constrained to what you give it, and styled to reflect your brand's personality.
• Crafter Q: A deep look at the new standalone SaaS product, its beta results (higher engagement, lower bounce rates, reduced support load), pricing tiers ($30 to $400/month), and the enterprise private-hosting option for teams with strict data residency needs.
• Owning the Conversation: Why letting ChatGPT or other public LLMs "converse about" your site means giving up control of the experience, the brand, and the backend analytics on what your visitors are actually asking.
• Browsing → Searching → Conversing: The evolution of how people find what they need online, and what the shift to conversation means for CMOs and digital strategists.
• MACH Compliance Without the Alliance: How Crafter approaches composable, API-first, microservices-based architecture without formal membership in industry alliances, and why that's a deliberate choice.
• Built-In Personalization and 60+ Integrations: Content targeting with persona support right out of the box, plus a marketplace of integrations for more advanced personalization and A/B testing needs.
• Three Ways to Get Started: Open source download, self-hosted enterprise, or SaaS cloud trial, and what the right fit looks like depending on your team's size and setup.

Mike's core argument is that AI readiness isn't a feature you bolt on. It's an architectural decision you make at the foundation. Whether you're evaluating your next CMS, accelerating a site migration, or figuring out how to put a branded conversational experience on your website, this episode gives you a clear and practical framework for thinking about all three.

🎧 Listen now and join the conversation!