Sitecore Stream delivers AI-driven generative intelligence purpose-built for marketers.
Sitecore Stream was announced at the DXP leader’s bi-annual Symposium onstage in Nashville in October 2024. In the opening keynote, Sitecore CEO Dave O’Flanagan and Chief Product Officer Roger Connolly unveiled a set of upcoming futuristic features which include brand-aware Al, Al-enhanced workflows, and generative co-pilots designed to unlock marketer developer productivity across a wide variety of activities.
If your brand is already using Sitecore’s digital experience solutions, Stream will offer AI-fueled marketing capabilities that enhance key stages of the content creation lifecycle. For XM Cloud or Content Hub users, Stream’s portal and embedded tools bring intelligence and efficiency to your content marketing.
Since pioneering the CMS solution in the early 2000s, Sitecore has remained a leader in content management and innovation, and is now shaping the future of generative content marketing. I wonder if this new category of software deserves its own acronym: could you classify Stream as a Generative Marketing System (GMS)?
I’ve used some similar products as a consumer, subscribing monthly to services like Jasper, Journalist.AI, and WriteSonic. Pencil AI and Writer are building disruptive momentum in the enterprise space. Sitecore’s Stream is distinct from these solutions in a few key ways. Brands building their digital experiences on the Sitecore platform have a large volume of brand assets and customer data that already live in their Sitecore instance. Your customer data gets stored in Sitecore’s CDP (Customer Data Platform), alongside audience segments and next-best-action decisioning canvases. Your corporate assets and collaborative content are organized in Sitecore’s Content Hub DAM (Digital Asset Manager) as a single source of truth for modular content spanning from logos, to product descriptions to brand imagery. The other solutions I’ve experimented with exist separately from any CMS and all of a brand’s content and data. They end up requiring a lot of copying and pasting when you’re building content. Stream, on the other hand, is connected to your DXP technology and fundamentally understands your brand tonality.
Marketers can upload a set of assets that represent your marketing efforts across channels and Stream ingests them in order to ground the generative model in your brand’s voice and tone. Grounding involves converting your brand assets into knowledge , much like ChatGPT understands the contents of Wikipedia and publicly indexable sites, and ultimately exposing this as an interface where any business user can intuitively generate intelligent content and experiences. There is a lot of value in Stream’s proximity and tight integration to the systems storing content and powering various websites, portals, kiosks, mobile apps or landing pages.
Why Sitecore Stream?
Marketing programs for a major brand at enterprise scale requires a great deal of content and complexity. A modern marketer faces ever-expanding roles and responsibilities. A recent Gartner report points to marketing budgets globally as a percentage of total revenue for major enterprises: at less than 8% marketing budgets are at a post-pandemic low. Your digital teams are being asked to do more with less while the number of channels and audience segments continue to expand.
How do you ensure consistency of an authentic brand identity and tone across dozens of channels and campaigns? If you’re deploying the same marketing tactics you used last year, you’re already behind the curve of innovation transforming the future of content operations.
First Impressions of the Stream Preview at Symposium
Sitecore Stream was announced in October 2024 and while it’s not generally available until early next year, Sitecore Symposium attendees saw early previews at the event. I watched the announcement from mainstage. I also attended a 45-minute, in-depth demonstration at the Solution Pavilion led by a couple of Sitecore’s stellar Solution Engineers.
Stream introduces a comprehensive set of tools and solutions focused on the work that marketers do. From ideation to activation, Stream intelligently addresses complexity and challenges of modern marketing for enterprise teams. It reimagines the content lifecycle with an AI-first approach.
Imagine this: you’re a retail bank that just published a lengthy article about the economic impact of changing interest rates. You prompt Stream to summarize the article in an email that would resonate with 3 distinct audience segments like Millennials, Financial Professionals, and Retirees. In seconds, Stream generates drafts ready for review and approval. These can be used to promote your new interest rate article to distinct email audience lists. You can even transform summaries by reframing them for use across different social platforms in a way that matches your brand standards and defined approach to each platform. Suddenly one article can be positioned for your key audience segments and optimized for promotion across your top marketing channels.
For more than a decade digital teams have been chasing an elusive promise of seamless 1:1 personalization. With cutting-edge AI capabilities like Stream your marketing team can get much closer to delivering precisely personalized experiences. They get the keys to a generative brand engine that does much than manual content creation in less time. Stream is driving closer to efficiently meeting content demands of an enterprise business by producing content at an increased velocity that can exceed customers’ expectations.
Sitecore groups key features into 3 broad categories.
1. Brand Aware AI: your new brand assistant
Brand-aware AI is like the creative brain at the center of your martech stack.
What does it mean for AI to understand your brand? Brands communicate in nuanced ways across different channels. Messaging on TikTok will probably fall flat on LinkedIn, and it’s important to balance short content that captures awareness and attention with longform content that builds credibility and nurtures the base.
Stream fundamentally understands your brand’s DNA by training on style guidelines and grounding its creative production approach in themes learned from high-performing articles and content. You’ll want to put some serious thought into the brand content feeding the system because we all know the old adage… garbage in, garbage out. If you stay true and authentic to your brand’s core values then Stream can help your business ideate and develop content strategies that move the needle with measurable growth.
Chat with the new Brand Assistant. Ask it questions. Prompt it to generate content or help you ideate new campaigns.
2. Generative Copilots: like Brief Assistant
Stream is focused on marketing jobs to be done. Stream can reduce the overhead for one of these jobs by solving a common and repetitive challenges for marketers – keeping content briefs on-brand.
With the upcoming SItecore Marketplace, I’m very interested in the ability to create generative Stream Agents. I built my own GPT last year using ChatGPT, and you can read more about that process here. Brief Assistant is one co-pilot or agent-type use case, but I’m sure there will be many others soon. Sitecore outlined the brief assistant capabilities in detail.
You can work with Stream’s brief assistant to brainstorm new campaigns with AI. The brief assistant is one of many upcoming generative copilots which Stream intelligently connects to to all of Sitecore’s SaaS products, and apparently XP based on a recent keynote nod. It was built based on feedback from multiple major enterprise marketing teams and crafted to simplify the complexity of campaigns in early stages of the content creation lifecycle.
Start with an idea. Generate a campaign plan.
3. AI Enhanced Workflows
The final category includes new intelligent features which are embedded directly into the UI of Sitecore’s existing SaaS solutions. New buttons or menu items appear to apply AI to your content. Here Stream aligns AI insights and generative technology to common specific activities and to your business goals.
Where do AI features of Sitecore products end and where does stream begin? The boundaries of Stream-versus-native-product are still a little blurry for me. Sitecore’s Content Hub, for example, boasts new AI-powered search capabilities as part of the core DAM functionality. The UI button exposing this task looked like it referenced Stream in the demo I watched. Do you need an additional license to use any new generative feature? This could be hard to maintain. Seemingly overnight applications that I use frequently are predicting sentences or offering intuitive chat-like interactions a product or brand. In Microsoft’s Outlook I can hit tab to complete a sentence, and Amazon just introduced a new chat agent called Rufus that helps you make purchasing decisions.
Liz Nelson, Sitecore XM Cloud’s Product Manager, uses the term “hesitant marketer” to describe someone that knows they need to begin personalizing and testing, but they’re not sure where or how to start. I’ve met this marketer dozens of times. They have great big ideas but overwhelmed by tactical day-to-day execution. They never quite get around to the advanced “Phase 2” digital marketing features that were the reason they purchased the software to begin with. By automating repetitive tasks, AI workflows can free creatives up to be more creative. They can suggest strategic tactics to help point marketers in the right direction for getting started.
I’ve outlined several use cases below.
A) AB Tests: Experimentation and testing in XM Cloud is a great example of a new Stream AI button that appears within the existing UI. As Mick Wingert, Sitecore’s head of Solution Engineering in Australia Pacific, explained in the Solution Pavilion, Stream guides a marketer through setting up an AB test with new AI recommendations. The recommendations appear on a notification pane almost like a tooltip for your existing content. For example, you can click the button on your Homepage, and it reveals a suggestion to start testing the Title of a Hero component. If you think Stream has a good idea you can go ahead and get started with the AB test.
B) Assisted component building: This one seems more like a longer-term roadmap item, but it’s hard to keep up with how fast the AI train is moving these days. Imagine a marketer describing a desired component in conversational language like, “I need a Promo component with the image on the left, a title/subtitle/button, with a blue background.” The system will potentially draft a generated reusable component which can be placed on multiple sites. Furthermore, one day soon Stream might begin to suggest alternative layouts and visual components for experiences.
C) Personalization suggestions: While technically XP’s existing Cortex functionality could recommend a personalized experiences, the feature was buried behind a long list of configuration steps. Stream makes this much easier. Since the brand-aware AI is trained on your key audience segments, it could make precise recommendations to tailor existing content for them at different funnel stages. This gets really interesting once Stream has access to the decisioning and audience segments which are building in Sitecore CDP.
D) Asset and item tagging: This use case seems like low-hanging fruit. Content Hub is packed with metadata that supports and describes images, text snippets, projects, and campaigns. Populating these granular tags can be a meticulous and repetitive task at scale-a perfect fit for AI. Stream can be extended to develop customer-specific semantic tagging techniques riding on Azure OpenAI that can reduce the repetitive burden of metadata tagging from the point of ingestion or creation.
Sitecore’s new market positioning refers to capabilities above as the Intelligent DXP. I’m honestly shocked at the staggering pace of innovation since Sitecore built and released this in less than a year. With Stream, marketers can leverage generative AI into the content lifecycle within the specific strategic context of their brand and their objectives.
Much like hybrid vehicles are powered by both gasoline and electric batteries, Sitecore Stream combines a set of human workflows and enhances them with AI. These workflows aren’t intended to cut the marketer out of processes, but to make her faster and more efficient. Hybrid workflows decrease the time it takes during content creation and review cycles by streamlining the experience creation process.
Governance and Trust
When introducing any new technology into your martech stack, keeping your brand secure and protecting your data are two top priorities. Trust is central to Sitecore Stream’s product strategy.
Sitecore Stream does not share one customer’s data with another. Product leaders repeatedly emphasized that “Sitecore will never use your brand or customer data to train its model for other customers.”
Stream keeps your data secure in a partitioned instance, and has built the tools with this premise from launch.
Composability as the Foundation for Intelligent Digital Experiences
The last major move Sitecore made was to decompose their extensive Experience Platform (XP) system which is a wide set of platform capabilities that support a variety of major marketing use case all-in-one. Sitecore released a set of composable, cloud-native Saas solutions that were designed for specific activities like managing website content, personalizing experiences, or creating and storing brand assets. Composability was (is?) an architectural martech paradigm that involves breaking down various pieces of digital marketing functionality into distinct connected solutions. The initial wave of this was tech-heavy and easier for developers and technologists to use than it was for their counterparts on the business side of marketing or products.
The investment in these foundational composable products paid off. Sitecore’s new SaaS solutions like XM Cloud (the current flagship Sitecore CMS) have onboarded hundreds of customers and continue adding iterative enhancements. The latest features have make it easier for a business user to leverage without any technical knowledge and they include page building, form creation, basic personalization, experimentation, and even no-code component builders for reuse across sites. While adoption of generative AI has surged and evolved rapidly in Sitecore’s customer’s industries, their R&D team rose to meet the demand by investing heavily and following-through with the delivery of intelligent DXP capabilities.An Intelligent DXP creates opportunity for early adopters by connecting generative systems directly to content, data, and experience delivery.
Stream is coming soon but it’s not available until 2025. What can my marketing team do to prepare? I’ll share more on how to prepare for a future of generative marketing in an upcoming post next month.
Leave a Reply