Key takeaways from SUGCON Europe 2026

It is hard to believe it has already been over two weeks since we attended and presented at SUGCON Europe 2026 in London.

As usual, SUGCON did not disappoint.

The strength of the conference has always been its balance: deep technical implementation insight, practical delivery advice, strategic thinking, and a clear view of where the Sitecore ecosystem is heading.

This year, one theme stood above everything else.

AI is no longer a future topic

AI is now fully embedded in the way teams work.

The conversation has moved on from whether teams should use AI, which models they should choose, or whether the tooling is mature enough. Most people are already using it across planning, development, testing, content operations, and delivery.

Whether teams are using GitHub Copilot, Claude Code, Cursor, Windsurf, MCP servers, custom skills, or internal tooling, the direction is clear: AI is now part of the software development lifecycle.

That means the competitive edge is no longer simply using AI. The edge is in how well we structure, govern, and orchestrate it.

The real shift is orchestration

The most interesting conversations at SUGCON were not about isolated AI use cases. They were about orchestration.

Simple AI workflows are useful, but they only take us so far. The bigger opportunity is creating connected workflows where tools, systems, content, and agents can work together as part of a broader operating model.

This is where SitecoreAI becomes particularly interesting. The extensibility of the platform gives teams several ways to approach this, from native capabilities in Sitecore Agentic Studio through to MCP servers, agent APIs, and external orchestration platforms.

A strong example came from Miguel Minoldo and Ramkumar Nambhi Krishnan Dhinakaran, demonstrating how SitecoreAI could be used with n8n as part of a broader translation workflow.

That kind of example matters because it moves the conversation away from AI as a feature and towards AI as infrastructure for content operations.

Content operations are being redefined

One of the most practical opportunities for Sitecore AI is augmenting content teams.

Most organisations are under pressure to produce more content, across more channels, in more languages, with the same or smaller teams. AI-enabled workflows can help address that pressure when they are designed properly.

Use cases such as autonomous content auditing, scaled translation, metadata enrichment, structured content review, and low-code workflow automation are no longer theoretical. They are becoming realistic delivery patterns.

The important point is that the goal is not to replace content teams. It is to increase the amount of meaningful work they can complete without increasing operational complexity.

Security and governance are now central

As we move further into an agentic world, security becomes more important, not less.

Sergei Baronov’s session on securing agents with fine-grained authorisation was one of the most important talks in this area. The key point was simple: agents should not be given broad permission to do whatever they want.

They need carefully designed boundaries.

That means thinking beyond traditional role-based access control and towards more granular, relationship-aware permission models. The more systems we connect, the more deliberate we need to be about what agents can see, change, publish, trigger, or escalate.

This connects directly to how we think about AI governance in delivery. Evals help us understand whether models are behaving as expected. Fine-grained permissions help ensure agents cannot take actions they should never be able to take in the first place.

Both matter. One reduces behavioural risk. The other reduces operational risk.

AI does not remove the fundamentals

One of the best things about SUGCON is that it does not lose sight of practical engineering problems.

Alongside the forward-looking AI content, there were still targeted sessions focused on real day-to-day delivery challenges.

Simon Hauck’s session on Sitecore AI publishing was a good example. Publishing content to Experience Edge is one thing. Making sure a Next.js application responds correctly, without unnecessarily clearing the entire site cache, is another.

The use of targeted revalidation tags is exactly the kind of practical implementation detail that makes a significant difference in production.

It was a useful reminder that AI does not replace good engineering. It increases the need for it.

Design systems are becoming part of the orchestration layer

Another interesting thread was the relationship between AI, design systems, and front-end delivery.

As more teams explore component generation and AI-assisted development, the quality of the underlying design system becomes critical. If we generate components without respecting design rules, accessibility standards, brand constraints, or established implementation patterns, we simply create a new kind of technical debt.

The opportunity is to connect tools like Figma, design systems, component libraries, and code generation into a more controlled workflow.

Again, the theme is orchestration.

GEO and AEO continue to matter

There was also continued focus on GEO and AEO: how generative engines and answer engines understand, interpret, and reuse website content.

Mark Lowe’s work in this space remains particularly valuable. His focus on how tools like ChatGPT and Gemini read Sitecore sites is a useful reminder that content structure now matters to both humans and machines.

Well-structured content, clean rendering, semantic markup, and clear information architecture are becoming increasingly important as AI-driven discovery evolves.

Upgrades are still a major opportunity

Alongside the new AI capabilities, modernisation remains a major priority across the Sitecore ecosystem.

Many organisations are still dealing with older XP or XM estates, heavy customisation, legacy integrations, and years of accumulated technical debt.

AI tooling can help reduce some of the pressure around upgrades by accelerating analysis, documentation, testing, and migration planning. But the hard part remains knowing what to keep, what to simplify, and where risk really sits.

That is where senior delivery experience still matters.

The two biggest themes: orchestration and security

If we had to reduce the conference to two themes, they would be orchestration and security.

Everyone now understands that AI can help teams move faster. The more important question is how we connect these tools into useful workflows without losing control.

That means designing systems where AI can support delivery, content operations, publishing, translation, optimisation, and development, while maintaining clear governance around permissions, quality, performance, and business risk.

This is the direction the industry is moving in, and it strongly reflects how we are thinking about our own roadmap at Think Fresh Digital.

What this means for us

For us, the most valuable takeaway was confirmation that our direction is the right one.

We are focused on helping organisations reduce risk in complex Sitecore change. AI does not change that. It expands what is possible, but it also raises the importance of architecture, governance, security, and delivery judgement.

The next phase of value will not come from adding AI into every process without thought. It will come from knowing where AI genuinely improves outcomes, where human expertise still needs to lead, and how to connect everything safely.

The community remains the strength of SUGCON

Beyond the sessions, the real strength of SUGCON is still the community.

We had conversations with people from the UK, Europe, Canada, and the United States, including many we have worked with before but had not seen in person for a long time.

Those hallway conversations are often where the most useful insights happen. They help test ideas, challenge assumptions, and shape what we do next.

We are looking forward to catching up on the sessions we missed once they are published online.

SUGCON remains one of our favourite conferences because it brings together technical depth, practical honesty, and a genuinely strong community.

This year gave us plenty to think about, and plenty to build on.