Episode #56: Scott Brinker on How AI Will Reshape Martech in 2025
AI is changing everything—but not the way you might think
Earlier this month, Scott Brinker and co-author Frans Riemersma released their latest report: Martech for 2025.
It’s 108 pages of dense insights on where Martech is headed—and as you might imagine, it’s largely focused on the core ways AI is re-shaping our discipline.
It was my pleasure to sit down with Scott to dig into the conclusions. Listen above for our full discussion, or read below where I summarize some of the key ideas that stood out to me.
We need to evaluate AI in the context of the Hype Cycle
We’ve all seen the incredible barrage of AI hype over the past few years—a.k.a., the “Peak of Inflated Expectations”.
And there’s no doubt that some some use cases are still immature or are likely to flame out into the Trough of Disillusionment (AI SDRs?).
However, if we take a slightly longer view, it becomes clear that many challenges with AI applications simply reflect teething issues. They will be outgrown as the technology matures and we ascend to the “Plateau of Productivity.”
We can argue about how AI will shape what’s to come, but there’s no denying it as a seismic shift in technology that will have profound impact.
The Five Segments of AI Innovation
The report highlights five key segments where AI innovation is taking place, each with different dynamics, advantages, and challenges.
Indie Tools:
Small-scale AI-powered tools, often without institutional funding. These tools (e.g., AI-powered note-takers, data scrapers) focus on automating specific point use cases and integrate with larger platforms.
Examples: Tools like tl;dv for meeting transcriptions or Headlime for copywriting.
These tools often complement rather than compete with incumbents.
Challenger Platforms:
New AI-native platforms seeking to disrupt major martech incumbents.
They have the advantage of being able to completely rethink their architecture and workflows from an AI-native perspective, without constraints on backwards compatibility.
However, they also face significant challenges; they lack the large established user bases of incumbents, who are often already deeply entrenched in tech stacks.
Examples: AI-native CRMs like Day, Clay for prospecting, or 11x for AI agents.
Incumbent Platforms:
Established players (e.g., Salesforce, Adobe, HubSpot) are embedding AI capabilities into their platforms, using both organic development and acquisitions to stay competitive.
Examples: Adobe’s GenStudio and Salesforce’s Generative Canvas show how incumbents are incorporating AI to provide end-to-end workflows.
Custom Apps:
AI will enable completely custom apps in a variety of ways. For example, using Retrieval-Augmented Generation (RAG), businesses can integrate proprietary data and logic into AI-powered workflows without traditional software development.
AI coding assistants can also make it faster and easier to build quality software internally.
Examples: Klarna replacing SaaS tools with custom AI solutions demonstrates the potential for organizations to reduce reliance on pre-built commercial software.
Service-as-a-Software:
AI agents are increasingly capable of performing tasks autonomously, substituting labor in areas like customer service or content distribution. This shift transforms traditional SaaS into “service-as-a-software.”
Instead of paying for software licenses, companies pay for successful outcomes, such as customer engagement or conversions.
Examples: Sierra.ai offers AI-powered customer service interactions billed on a pay-per-resolution model.
A robust data layer is the foundation of your AI Strategy
To be fully integrated with a company’s workflows, AI needs access to a wide variety of organizational data. As the report notes, “the key to this is a universal data layer that aggregates data from all the different applications in your stack and makes it available for any other app to use.”1
And while the report found that 70% of respondents have some sort of data warehouse consolidating this data, it also identifies an organizational role for “Big Ops,” a team that needs to manage the many complex ways that applications, automations, and AI are interacting with company data.
From personal experience, I’ve noted that this sort of rigorous data strategy and management is often absent, especially in smaller companies. In that case, there’s room for ops teams to step into the breach to strengthen the data foundation and ensure AI apps are enabled to deliver value.
Agents and Large Action Models are key to operationalizing AI
The report opened my eyes to how transformational AI agents could potentially be. I realized I hadn’t understood their full implications.
ChatGPT defines an AI Agent as “a system or program that autonomously perceives its environment, processes the information it receives, and makes decisions or performs actions to achieve specific goals.”
Rather than using AI to perform tasks via chat interaction or as a step within a rules-based rigid workflow, an AI agent can in some sense form the center of a workflow, taking information and then manipulating other systems to achieve an outcome.
For example, Scott and I discussed a hypothetical campaign operations agent that could chat with a marketer then build email campaigns in a marketing operations platform using APIs or UI manipulation.
Thanks to Our Sponsor
Many thanks to the sponsor of this episode - Knak.
Knak is a modern, slick, and beautiful email and landing page builder that integrates directly with your marketing automation platform.
Click below to learn more and get a special offer just for RevOps FM listeners.
About Today's Guest
Scott Brinker is VP Platform Ecosystem at HubSpot and previously the co-founder and CTO of ion interactive, a SaaS company that pioneered interactive content for global enterprises and was acquired in 2017.
Since 2008, he’s also run the Chief Marketing Technologist blog, chiefmartec.com, with over 50,000 readers, and creator of the Marketing Technology Landscape, mapping the growth of the marketing technology industry from a few hundred vendors to over 14,000.
He wrote the best-selling book "Hacking Marketing," published by Wiley in 2016, and co-authored of the article "The Rise of the Chief Marketing Technologist" published in Harvard Business Review. He is a frequent keynote speaker at conferences around the world on topics of marketing technology and agile marketing.
Key Topics
[01:26] - Main take-aways from the report
[03:31] - How AI can lead to more differentiated marketing
[06:07] - Efficiency vs. effectiveness from using AI
[11:33] - AI and the Hype Cycle
[15:28] - Innovator’s Dilemma and compressed innovation
[17:16] - Segments of AI innovation
[21:06] - Innovation challenges fo r legacy incumbents
[22:56] - Last-mile issues with AI feature quality
[29:10] - AI agents
[40:54] - Orchestration layer
Resources
Page 30