Sitecore CEO: Marketing, content, experience AI prime focus

Since last year's conversation with Eric Stine, the Sitecore CEO has overseen the company's pivot from generative AI to what it calls the "agentic web," where both prebuilt and homegrown agents can plan, execute and optimize complex marketing tasks across channels.

This wide-ranging Q&A delves into how agentic AI is reshaping the technology universe, particularly large digital experiences. Also: Stine lays out his theory that AI is not an existential threat to anyone or any company -- it's just the latest wrinkle in a long set of enterprise technology advancements.

Editor's note: This Q&A was edited for clarity and brevity.

Where are Sitecore users seeing the earliest actual ROI from agentic AI, rather than aspirational goals?

Eric Stine

Stine: The marketer is changing because the market itself has changed in response to agentic AI from the outside. It has changed consumer behavior, customer behavior, the behavior of the market, and how perceptions are formed in the market -- whether it is changing how consumer goods are found now in Gemini summaries and what pops up next to them, as opposed to the ranking of sponsored listings, to the ability to go into something like ChatGPT or Perplexity and compare financial products or healthcare providers. You can use something like Claude to generate and vet an RFP process in B2B.

As a result, we are really focused on three basic components of what the marketer does, and we are seeing adoption of agentic workflows in three areas: discovery, connection and conversion.

The internet is being rewritten for machines. Machines will interpret it, but humans will still experience it -- so the last mile still belongs to the marketer. Customers have to land in the right place with something relevant to them that supports the decision process that was consolidated into an LLM for them. And all that needs to be orchestrated.

So, we see localization, translation, account-based marketing and context-aware content all happening in Sitecore. AI is to facilitate the processes of discovery, connection and conversion, orchestrated at scale.

As a SaaS company, are you prepping your bunkers for the SaaSpocalypse?

Stine: Reports of the death of software have been prematurely written. We've been through this market cycle before, 12-13-14 years ago, when SaaS started replacing software at scale.

I think that from an enterprise technology perspective, much like the [cloud hyperscalers such as] Google Cloud Platform, AWS and Azure moved a range of technology surfaces off-premises. We will see traditional players like ServiceNow and Snowflake, traditional cloud providers like Microsoft and AWS, and a new generation of cloud providers, such as Claude and ChatGPT, become general-purpose stacks around which enterprise architectures are built.

The software players that survive will be the ones that have the domain expertise around how a process can be efficiently performed at scale and embed AI into those workflows.

I don't think that marketers, CIOs or anybody else in your business wants to vibe-code an edge case for account-based marketing any more than they want to vibe-code an edge case around employee relations, supply chain management or finance. Enterprise systems will still be necessary to create the process orchestration at scale that allows most companies to survive. It'll change for sure, but software will still be necessary.

A group of us at Informa TechTarget have dug in our heels and proclaimed the SaaSpocalypse a total mirage, and that's the hill we will die on.

Stine: Here's what I love about it, though: I started in the industry in 2001. I started in enterprise ERP and was in the industry during the major wave of ERP adoption beyond traditional manufacturing industries. I did CPG [consumer packaged goods] and retail. I worked in the public sector, regulated industries and financial services. I've been in the industry through the SaaS migration.

[A Bain & Company analysis] showed that enterprise software, as an industry, added sales capacity at a year-over-year increase rate of anywhere from about 11% to 23% for nearly a decade. Three years ago, it stopped. It didn't slow down. It didn't decelerate. It stopped for the past three years -- as an industry, we have added sales capacity at a rate of about 0% to negative 1% or 2%, year over year.

The reason? We had enough sales capacity for all the products that existed in the marketplace at that point in time.

What's great about this moment for SaaS companies is that it presents an opportunity for reinvention. We rewrote the world from hardware to software. We rewrote the world from software to SaaS, and now we're rewriting the software for AI. The whole reason we decided to include all our agents and workflows in our product, the reason we don't sell AI separately, is that our SaaS product is an AI product.

Software is automation. I believe our product isn't a product if it doesn't have AI in it.

As you plot Sitecore's roadmap, do you see agentic AI as a replacement for parts of the enterprise IT stack, the orchestration layer, or a little bit of both?

Stine: I always start with, "What does the marketer want to do, and what can the marketer do effectively?" Marketers tend to understand who their audience is, why they buy their product, and what makes their product great.

We want to take all the work out of the work, all the "Help me figure out from my customer base, 'Who is this campaign targeted for? How should it be positioned? What does it need to say? Can you find me governed assets for it? Can I modify this asset to meet the needs of the market?'"

I've had this feeling ever since, really, the pandemic ended -- it never really ended, it just sort of [slowly faded] away and left us with the pilot light of anxiety burning inside -- that we haven't caught up, that we're still behind. I feel that in any given moment, there are a thousand things I could be doing.

I think that there is so much potential in agentic AI to make it easier for people to work to their strengths. Because why would you ask somebody to work in ways that make them uncomfortable, that don't make them feel empowered and accomplished at what they're doing?

How are you thinking about pricing as agents drive usage beyond seat-based models?

Stine: We got rid of the seat-based model a year ago when I became CEO. We price based on, effectively, content interactions -- website visits and accessing pieces of content. The hope is that, whether it's a human or a machine, if you have compelling content that's being retrieved by its audience, we're adding value for you.

If we've helped you create something that somebody -- human or machine -- wants to read and make a decision based on it, then we're delivering value to you. So, it's really close to a utilization or consumption-based model.

Does it feel like this agentic and generative AI revolution is magnitudes more disruptive -- and shakes the foundation a lot more -- than SaaS or cloud enterprise tech ever did?

Stine: I want to say yes, because that's the expected answer, but my honest answer is no, it just isn't.

About a year ago, I heard a leader whom I deeply, deeply respect speak about the financial services industry. She was talking about how the adoption of AI is happening, you know, at five times the pace that SaaS adoption happened. That was five times the pace at which software adoption happened.

You have to move with the times, but you also have to recognize that disruption is cyclical.

Do I think things are happening faster? Do I think we're making the world both easier and more complicated at the same time? Yeah. I think AI is part of a continuum of technology disruption that people said about the factory, the airplane, the car, the television, the personal computer, the internet and social media.

Just because I believe that doesn't change my obligation to recognize that the technology is transformative and to make it the platform for the next generation of solutions. Saying no to your question doesn't mean I'm going to go out there and try to sell DOS-based or C++-coded on-premises software. You have to move with the times, but you also have to recognize that disruption is cyclical.

Don Fluckinger is a senior news writer for Informa TechTarget. He covers customer experience, digital experience management and end-user computing. Got a tip? Email him.