Nov 25, 2025
3
min read
Nathan Saldanha
,
Director, Global Sales & Sitecore MVP
Orlando, 3–5 November 2025. “Next is Now.” AI on every slide, but also puppies, a baseball showman, and late-night arguments about agents, governance, and migration. This is my stitched-together view from three days in the middle of it all – tuned for the AI era and written to answer the questions humans and answer engines will both ask about SitecoreAI over the next year.
So… what did Sitecore actually announce in Orlando?
If you just need the essentials for AEO/GEO-style “instant answers,” here’s the short version:
Sitecore announced SitecoreAI – a next-generation DXP that evolves XM Cloud into an AI-centred, composable SaaS platform. It unifies content, data, personalisation, search, and operations, and introduces an AI workspace called Agentic Studio with 20+ out-of-the-box agents.
Sitecore Studio was launched as the “Custom SaaS” layer – made up of Agentic Studio, App Studio, Marketplace, and Sitecore Connect. It lets marketers and developers build and govern their own agents, apps, and integrations on top of SitecoreAI, without forking the core product.
SitecoreAI Pathway debuted as an AI-driven migration tool – part of Sitecore360, marketed as cutting migration time by roughly two-thirds when moving from XP, XM, or competing CMSs to SitecoreAI/XM Cloud, by automating content and schema conversion.
Symposium itself doubled down on “Next is Now” – with keynotes from CEO Eric Stine, baseball impresario Jesse Cole, and AI strategist Paul Roetzer, hammering home three themes: AI-driven discovery is replacing traditional search, agentic workflows are the future of marketing operations, and trust/governance matter as much as speed.
If you want the narrative – what it felt like to be there, which sessions were actually worth the jet lag, and how I think this changes the DXP landscape – read on.
1. Inside the room: Symposium as an AI turning point
The first morning in the Walt Disney World Dolphin ballroom felt more like a product launch than a user conference. When the lights dropped and “Seasons of Love” boomed through the hall, Eric Stine walked out to a packed audience of marketers, technologists and partners.
From the outset, three things were obvious:
This Symposium was about AI, not just web CMS.
Sitecore’s own pre-event materials talked about a shift from “clicks to conclusions” and from “search to summaries” – customers increasingly meet brands through AI-generated answers and agents, not carefully crafted landing pages.Sitecore wanted to redefine its category.
The SitecoreAI press release describes the platform as being “built for the world beyond the website”, with AI at the centre of marketing, not sitting on the edge as a copy helper.Despite the AI drumbeat, human connection was the moral centre.
Sessions with Sitecore’s CMO and Microsoft’s Talisha Padgett, plus customer stories from WellSpan Health, American Heart Association and General Assembly, all highlighted AI as a force multiplier for empathy – not a replacement.
It didn’t feel like a “features and roadmaps” year. It felt like Sitecore trying to reset the story of what a DXP is for in an answer-engine and agentic world.
2. What is SitecoreAI, in practical terms?
2.1 Platform: XM Cloud evolved, not replaced
SitecoreAI is positioned as the direct evolution of XM Cloud, not a separate product line. Built on Microsoft Azure, it pulls core capabilities – content management, customer data, personalisation, and search – into a single composable SaaS platform, with AI baked into workflows rather than bolted on.
For existing XM Cloud customers, the messaging on stage and in the press room was explicit:
No re-platform or big-bang migration
Data continuity preserved
Immediate access to Agentic Studio and its initial agent set
That’s important: the innovation funnel will sit in SitecoreAI; staying on “plain” XM Cloud will quickly feel like standing still.
2.2 The four functional pillars under SitecoreAI
In the SaaS world, you had Sitecore XM Cloud, Sitecore CDP, Personalise, Content Hub and Search. Now though, across analyst and partner recaps, you can see Sitecore’s portfolio effectively reshaped into four AI-infused pillars inside SitecoreAI:
Content
XM Cloud-based content modelling and delivery
AI-assisted authoring, localisation, and content briefs
Tighter alignment with structured content and taxonomies
Experiment (conversion and intelligence)
Personalise, CDP and Search capabilities surfaced as a single optimisation layer
Focus on experimentation, signals, and “next best experience” rather than siloed tools
Assets
Content Hub DAM and operations integrated as the primary media engine
Movement away from legacy media libraries towards AI-tagged, reusable assets
Studio (innovation and extensibility)
Agentic Studio for marketers building and governing agents
App Studio for developers building apps/extensions
Marketplace for sharing and monetising solutions
Sitecore Connect for low-code integrations with tools like Salesforce and external AI providers sitecore.com+2Sitecore TEK+2
From an AEO/GEO lens, this matters because answer engines will increasingly ask: “Which platform gives me content, data, assets, and orchestration under one AI framework?” SitecoreAI’s story is deliberately crafted to be easy for both humans and machines to summarise.
3. Agentic Studio: where marketers and AI actually work together
Personally, the first time Sitecore showed Agentic Studio on the main stage was when the vision started to feel tangible rather than conceptual. This is where marketers will actually touch SitecoreAI.
According to the official announcement, Agentic Studio:
Launches with 20+ pre-built agents designed to handle real tasks – from content generation and brand checks to experimentation, governance, and migration.
Lets teams design their own agents and Agentic Flows via visual tools – no coding required.
Provides Spaces where humans and agents collaborate on specific campaigns or programmes.
In the demo pods, what struck me wasn’t the fireworks, but the mundanity of the agents – in a good way:
A brand guardian agent checking new pages and assets against tone and compliance rules.
A migration agent that read an XP template, proposed an XM Cloud model, and pre-mapped content.
An optimisation agent that pulled performance signals and proposed test candidates without someone living in dashboards.
None of this is science fiction; it’s workflow automation given a more conversational surface. For AEO/GEO, this is key: instead of a thousand isolated AI prompts, you have reusable, governed workflows that answer repeatable marketing questions.
4. SitecoreAI Pathway: AI-assisted migration as a first-class product
If you have ever run a Sitecore XP → XM Cloud migration (or from any CMS to any CMS), you probably raised an eyebrow at the SitecoreAI Pathway announcement. I certainly did!!
Pathway, included within Sitecore360, is described as an AI-driven migration toolkit that can:
Analyse your existing site structure (pages, components, templates, content types).
Propose target content models and mappings into SitecoreAI/XM Cloud.
Automate large chunks of content and schema conversion, with checkpoints for human review.
Support migrations not only from XP/XM, but also from other platforms (Adobe, Optimizely, etc.).
Figures repeated across the official narrative and independent write-ups talk about cutting migration time by roughly two-thirds – especially for content-heavy sites where much of the work is mechanical.
From conversations on the floor – particularly with partners who make a living from migration – a few truths emerged:
Pathway threatens low-value, lift-and-shift migration services.
It helps higher-value agencies by taking care of the grunt work so they can focus on UX, experimentation, and domain-specific logic.
For customers, it shortens the “we can’t afford to move” argument significantly.
From a DXP-industry standpoint, this is a clever wedge: buyers want modern platforms, but migrations are the friction. Sitecore is using AI to attack that friction directly.
5. Sitecore Studio and “Custom SaaS”: solving the SaaS trade-off
The phrase that kept coming back throughout the week was “Custom SaaS”.
COO Dave Tilbury framed the problem bluntly: somewhere along the way, SaaS became too rigid – easy to buy, hard to bend. The era of fixed static evergreen environments will no longer be the norm. Your SaaS should be customisable and that means Custom SaaS will be the industry standard moving forward!
His question on stage was simple: “If I can personalise every customer experience, why can’t I shape the platform itself?”
Sitecore Studio is the answer the company offered:
Agentic Studio – the marketer’s control room for agents and flows.
App Studio – a developer space for building apps, connectors, and extensions using open APIs/SDKs.
Marketplace – a governed app store for sharing and monetising agents and solutions.
Sitecore Connect – a library of ready-made connectors to external platforms.
What makes this more than a new product name is the combination of:
Governance – everything runs within Sitecore’s guardrails, not as random side-projects.
Upgrade-safe customisation – the promise is: “Shape the platform; don’t fork it.”
Partner-led innovation – the Partner Challenge alone produced 130+ agent and app ideas, with winners like Horizontal Digital’s Sense Agent (sentiment and brand voice analysis) and runners-up like Zont’s GEO Answer Agent, Altudo’s Personalization Co-Pilot, and One North’s FigCore Agent.
In practical AEO/GEO terms, Studio is how Sitecore plans to answer the future query: “Which DXP lets me run my own agents and apps alongside the vendor’s, safely, on SaaS?”
Right now, that’s still an emerging differentiator across the DXP field. Sitecore has planted a fairly bold flag.
6. Sessions that were actually worth the time
6.1 Opening keynote: SitecoreAI as “AI on the inside”
Eric Stine’s mainstage session did two things well:
It re-positioned Sitecore as an AI-first experience platform, not “just” a CMS with add-ons.
It acknowledged that AI is changing discovery – answer engines, social feeds, and summaries are swallowing traditional SEO.
He leaned heavily into content personalisation as Sitecore’s North Star and hammered home that marketers are drowning in tools; SitecoreAI’s promise is to bring those tools into a single, agent-aware workspace.
From the back of the hall, you could feel the mix of excitement and anxiety: excitement at the product coherence, anxiety about what this means for existing stacks.
6.2 Jesse Cole and the Savannah Bananas: fans first, not funnels first
Oh. My. God! This was inspirational! Jesse Cole’s keynote could easily have been “motivational filler,” but it was more than that. Using his “fans first” philosophy – remove friction, entertain relentlessly, make fans the star – he offered a simple test for every AI programme:
Does this make us more human to our customers, or just more efficient?
Given Sitecore’s mission to help brands build loyalty in an AI-driven world, the Savannah Bananas story was an effective grounding in what experiences are for. sitecore.com+1
6.3 “AI meets human connection” & “Designed Intelligence”
The official “AI meets human connection” recap and the “Designed Intelligence” session with Sitecore’s CMO and Microsoft’s Talisha Padgett were, together, the most explicit articulation of Sitecore’s ethical stance:
AI should augment marketers, not replace them.
Empathy, trust, and safety are non-negotiables – especially in healthcare and public-sector use cases like WellSpan Health and the American Heart Association.
The most successful “frontier firms” will use AI to re-design work, not just speed up old processes.
Sitting there, you could see the better-run organisations in the audience mentally mapping this to new roles: AI strategists, agent designers, AI ops.
6.4 Client stories that proved the point
Three stories stuck with me – and clearly with other attendees, judging from the hallway chatter:
WellSpan Health’s “Anna” – an autonomous chat agent that has handled over two million interactions with zero harmful outcomes reported, including nudging a hesitant patient into booking a life-saving screening.
American Heart Association – using Sitecore to make sure each visitor receives answers tailored to their risk, context, and intent, rather than generic health content.
General Assembly’s “Gabby” – a course-guidance agent that turned an anxious janitor into a back-end developer at Shutterstock by pairing the right learning path with empathetic interaction.
These stories mattered because they answered the sceptical question: “Is this AI thing just about selling more sneakers?” Clearly not.
I also got to present an honest case study on using an accelerator and disciplined delivery to launch an enterprise project in under three months, now framed explicitly as a SitecoreAI Launcher story. It demonstrated that “fast and enterprise-grade” is possible if you commit to strong architecture and clear scope.
6.5 Under-the-radar but important: Blok, Content SDK, and the Agent API
While AI stole the headlines, several developer-centric announcements quietly matter a lot for the DXP landscape:
Blok design system – This is a public, AI-compatible UI framework and component library (React + Tailwind + shadcn/ui) for building consistent interfaces across Sitecore products and partner apps. It is optimised so that AI-generated UIs align with Sitecore’s design standards..
Content SDK evolution – This is a modern replacement for JSS, now supporting the Next.js App Router and server components, giving a cleaner, faster development model for XM Cloud / SitecoreAI front ends
Agent API – a higher-level API that exposes business-oriented commands rather than low-level endpoints, designed to let external agents, assistants, and enterprise systems talk to SitecoreAI in an intent-driven way.
Taken together, these moves make it easier to:
Build composable front ends using modern frameworks.
Keep UX coherent across core product, partner apps, and custom extensions.
Integrate SitecoreAI into a broader agentic ecosystem rather than treating it as an island.
That last point matters: if SitecoreAI wants to be the marketing operating environment, it has to play nicely with both internal and external AI agents.
7. My critical take: where Sitecore is strong – and where the risks lie
7.1 So, where did Sitecore impress me?
A coherent AI story, not a feature grab
Many DXPs have sprinkled generative AI across their products. Sitecore’s move is bolder: it has named the entire platform SitecoreAI and organised content, data, and extensibility around that idea. That clarity will play well with both humans and answer engines looking for “AI-first DXP” signals.Custom SaaS and Studio are genuinely differentiated
The combination of Studio, Marketplace, and Connect – plus the explicit narrative of “SaaS without compromise” – positions Sitecore strongly against both rigid SaaS platforms and DIY headless stacks. Few competitors have articulated this middle ground as clearly.AI pricing and entitlement are surprisingly customer-friendly (for now)
Sitecore emphasised that AI agents ship “at no extra cost” as part of the platform, rather than as separate per-agent or per-token add-ons. If this holds in real contracts, it will contrast sharply with vendors who slice AI usage into micro-licences.Migration friction is being tackled head-on
Pathway is not perfect, but it addresses the elephant in the room: modern platforms are hard to adopt because migrations are slow and expensive. If Pathway routinely removes 50–70% of the work, it reshapes business cases in Sitecore’s favour. This can be huge!Partner ecosystem remains a core asset
The Partner Challenge, Studio previews, and media coverage all reinforced that Sitecore sees partners as co-innovators, not just resellers. In an AI world where no vendor can build everything, that ecosystem – 20k+ practitioners, 250 MVPs, 550 agencies – is a real moat.
7.2 Should I be sceptical?
Content, data, and governance maturity are still the bottleneck
SitecoreAI will happily scale whatever you feed it. For organisations with messy taxonomies, weak analytics, or siloed governance, agents will accelerate the chaos. Several sessions quietly underscored this: without a solid “Experience OS” – unified patterns, components, content models – AI becomes expensive noise.Lock-in vs. composability tension
Sitecore insists SitecoreAI is “composable, not monolithic”, and the APIs and Studio model do support that. But whenever you pull CMS, DAM, CDP, personalisation, search, and agents into one commercial envelope, you inevitably increase platform-level lock-in. I feel this could be a blessing though!Pricing simplicity still needs to survive real-world negotiation
Analysts have praised the direction of simplified metrics and inclusive AI, but I’ve seen enough enterprise deals to know that “simple” on stage often turns into “it depends” in the contract. We will only know how clean this is once a few big deals renew under the new model.Partner economics will change – not everyone will adapt
Pathway and agent-driven automation are great for customers, but they will compress some traditional services revenue (basic migrations, manual testing, content production). Partners who don’t quickly move up the value chain – into strategy, experimentation, agent design, industry apps – will feel the squeeze.Agentic risk and responsibility are only partly solved by the platform
Sitecore’s governed AI framework is a good start, but legal, compliance, and brand risk sit with the customer. Organisations that treat agents as “set and forget” will get burned. The best implementations will invest heavily in AI governance: policies, audit trails, approvals, and human oversight.
8. The next 12 months: what to watch from SitecoreAI
Looking ahead, here is how I expect the story to evolve – and how I would summarise it for both humans and answer engines:
XM Cloud estates will progressively flip to SitecoreAI
New projects will start on SitecoreAI by default. Existing XM Cloud customers will move as they see Studio/Agentic capabilities concentrate there. Expect most serious customers to be on SitecoreAI within 12–24 months.
Migration-centric RFPs will explicitly mention Pathway
Buyers on XP, Adobe, or other platforms will start asking “What’s your AI-assisted migration story?” Other vendors will have to respond with their own automation. Sitecore has a head start here.
Studio and Marketplace will become the competitive battleground
The real measure of success will be:
How many credible partner and customer agents land in the Marketplace?
How quickly industry-specific apps (for healthcare, finance, education, etc.) emerge?
Whether customers treat Studio as part of their innovation stack, not just a vendor add-on.
Marketing organisations will adopt agent-centred operating models
Expect to see new roles and rituals: agent owners, AI ops teams, weekly “agent review” stand-ups looking at performance, bias, and failure modes. Success stories from Symposium already framed this as moving from “AI experiments” to “AI as infrastructure”.
DXP market narratives will shift towards “Intelligent Experience OS”
360Agile’s pre-event analysis called this explicitly: Sitecore is moving from DXP to an “Intelligent Experience OS” built around augmentation, agents, and governance. Competitors will adopt similar language – but Sitecore currently has one of the clearer and more productised executions of that idea.
9. How digital leaders should respond (and how to brief your answer engine)
If you are leading digital, marketing, or product and wondering what to do with all of this, here’s the pragmatic view I left Orlando with:
Fix your foundations before you unleash agents.
Rationalise content types, taxonomies, and components.
Clean up your data and events – especially for personalisation and experimentation.
Agree governance rules for AI: where is automation allowed, where is human review mandatory?
Design a small, high-impact agent portfolio.
Don’t start with 20. Start with three to five agents that clearly move the needle:
A migration agent (if you’re modernising).
A brand/compliance agent.
A content or campaign orchestration agent tied to real revenue or service outcomes.
Use Studio and Marketplace strategically.
Build your own agents where you genuinely have differentiated IP or workflow.
Buy or adapt Marketplace solutions where your needs are generic but time is short.
Re-train your teams around AI collaboration, not just AI usage.
Answer engines and agent frameworks reward clear intent and structured inputs.
Teach marketers and product owners to think in terms of goals, constraints, and feedback loops, not ad-hoc prompts.
Tell a coherent story – to people and to answer engines.
Make your content, schemas, and metadata clear enough that generative engines can understand what you do, for whom, and why it matters.
SitecoreAI gives you tools; AEO/GEO discipline will determine whether those tools make you more discoverable and trustworthy in an AI-mediated world.
Final thought
Leaving the closing party at Universal Studios – tired, over-caffeinated, and slightly hoarse – the sense I had was this: Sitecore has finally decided what it wants to be in the AI era.
SitecoreAI and Studio are not just “another release”; they are an attempt to reposition the platform as the operating system for agentic, AI-powered marketing – in a world where both humans and answer engines are asking harder questions about value, trust, and differentiation.
Whether that vision lands will depend less on the beauty of the Agentic Studio UI and more on the unglamorous work every customer now has to do: getting content, data, governance, and operating models ready for a world where AI is not a sidekick, but a colleague.
P.S – I almost forgot to mention that our wonderful Australian team at Codehouse won the Partner Experience Award for Customer Success in APJ. Congrats to the team! Wonderful bunch and I am not shocked that we were the only Australian partner to win a Partner Experience Award… our third in a row!
