As the digital landscape shifts from manual content creation to automated task execution, Eric Stine, the CEO of Sitecore, stands at the forefront of this evolution. With over two decades of experience in enterprise technology—spanning the major waves of ERP adoption and the SaaS migration—Stine offers a grounded perspective on the current AI craze. He views the rise of the “agentic web” not as a “SaaSpocalypse,” but as a natural progression in a long cycle of technological disruption. In this conversation, he explores how AI is rewriting the rules of marketing discovery, why the traditional seat-based business model is becoming obsolete, and how leaders can navigate a world where machines interpret the web before humans ever see it.
Digital strategies are shifting from simple content generation to an “agentic web” where tools plan and optimize tasks. How does this change the day-to-day workflow for a marketing team, and what specific metrics should leaders track to prove this shift is actually delivering real value?
The day-to-day workflow is shifting from manual execution to high-level orchestration, moving the focus toward three primary pillars: discovery, connection, and conversion. Instead of spending hours on the “work about work,” teams are now managing agents that can vet B2B RFPs or optimize how products appear in Gemini or ChatGPT summaries. To prove value, leaders should look beyond traditional vanity metrics and focus on how effectively their content supports the decision process that has been consolidated into an LLM for the consumer. We track how well these orchestrated workflows handle scale—such as the speed and accuracy of localization and translation—ensuring that the machine-interpreted web still leads to a meaningful human experience.
AI-generated summaries and LLMs are increasingly consolidating the discovery process for consumers. Since machines now interpret the web before humans see it, how do marketers ensure they still own the “last mile” of the experience?
While the internet is being rewritten for machines to interpret, the “last mile” remains the exclusive domain of the marketer because humans are still the ones having the final experience. Marketers must ensure that when a consumer moves from an AI summary to a brand property, they land in a context-aware environment that feels relevant and supportive of their intent. This involves a rigorous focus on “governed assets” and ensuring that the data being fed to LLMs is accurate and compelling. By mastering the transition from a machine-summarized discovery to a high-conversion landing page, brands maintain their relevance even when they aren’t the first point of contact.
Some argue that AI might make traditional enterprise software obsolete, yet others view it as just the latest technological cycle. Why is embedding AI into the core workflow now a requirement for a product to even exist, and what are the risks for companies that treat it as a separate add-on?
I believe a software product isn’t truly a product anymore if it doesn’t have AI embedded directly into its DNA; it is the modern equivalent of automation. Treating AI as a separate add-on creates friction and fails to address the domain expertise required to perform complex processes at scale. The risk for companies that don’t integrate AI is that they become “vibe-coders,” trying to patch together edge cases for things like account-based marketing or supply chain management without a robust framework. Traditional players will survive only if they embed these agents into their core workflows, effectively rewriting their SaaS offerings for the AI era.
Enterprise software sales capacity growth has leveled off or turned negative across the industry over the last few years. How does this stagnation force a reinvention of the SaaS business model, and what internal changes must a company make to transition from selling seats to selling outcomes?
For nearly a decade, the industry added sales capacity at rates between 11% and 23%, but that growth has hit a wall, hovering around 0% to -2% recently. This stagnation signals that the market has enough sales capacity for existing products, forcing a mandatory reinvention where we move from selling “seats” to selling genuine utility. Internally, this means companies must shift their focus toward building specialized agents that can handle the heavy lifting of process orchestration. At Sitecore, we’ve leaned into this by including all our agents and workflows within the base product, because we are no longer just a software provider—we are an outcome provider.
As automation reduces the need for human “seats” in software, pricing models are moving toward content interactions and utilization. What are the practical trade-offs of a consumption-based model, and how can organizations budget effectively when costs are tied to engagement rather than headcount?
Transitioning to a consumption-based model, as we did a year ago, means pricing is now tied to website visits and content access rather than the number of employees using the tool. The primary trade-off is a move toward a “utilization” mindset, where the cost reflects the actual value delivered—if a machine or human finds your content compelling enough to make a decision, we charge for that value. For organizations to budget effectively, they need to stop thinking about headcount and start forecasting their audience engagement and content lifecycle. It creates a more honest relationship between the vendor and the client, as our success is directly linked to the volume of meaningful interactions the client generates.
Marketing tasks like localization, translation, and account-based marketing are becoming increasingly automated through orchestrated workflows. How can teams balance this high-scale automation with the need for “governed assets,” and what is the best way to handle unique edge cases that the AI might miss?
The balance is achieved by using agentic AI to handle the scale of localization and translation while keeping the marketer in control of the strategy and the “governed assets.” We don’t want marketers to be stuck manually modifying assets for every minor market shift; we want the AI to handle those variations within a set of brand-approved rules. When unique edge cases arise, the system acts as a support layer that flags these instances for human intervention rather than forcing a person to “vibe-code” a solution from scratch. This allows the team to maintain brand integrity across millions of touchpoints without losing the human touch where it matters most.
Many professionals feel they are constantly falling behind despite having more tools at their disposal. How can agentic AI specifically reduce the “work about work” to let employees focus on their strengths, and what does a successful implementation look like for a high-pressure marketing department?
Agentic AI targets the “pilot light of anxiety” by taking over the mundane tasks of figuring out campaign targets, positioning, and asset retrieval. A successful implementation looks like a department where a marketer can ask an agent to find and modify assets for a specific campaign, freeing them to focus on the creative strategy they are actually passionate about. It’s about creating a workspace where people aren’t forced into roles that make them uncomfortable or unempowered. By automating the logistical overhead, we allow the marketing team to act as high-level strategists rather than data entry clerks.
What is your forecast for the agentic web?
I believe we are entering a phase where agentic AI becomes the general-purpose stack for the entire enterprise architecture, much like the cloud hyperscalers did a decade ago. While some see this as a radical disruption, I forecast that it will follow the cyclical pattern of the internet, the car, and the PC—it will make the world both easier and more complicated simultaneously. We will see a consolidation where a few major platforms provide the “brain,” but specialized software like ours will provide the “arms and legs” to execute domain-specific tasks. Ultimately, the companies that thrive will be those that recognize this shift is transformative but also realize that disruption is a constant rhythm in technology.
