MarTech Podcast ™ // Marketing + Technology = Business Growth Scott Brinker’s 2026 Martech Predictions Unpacked
Jan 5, 2026
Scott Brinker, VP of Platform Ecosystem at HubSpot and founder of chiefmartec.com, discusses the rapid evolution of the MarTech landscape. He reveals how AI has added 1,200 new vendors while changing traditional sales playbooks, empowering customers through agentic AI. Brinker introduces context engineering as the next step in AI deployment, emphasizes the need for better data layers, and contrasts deterministic automation with generative AI. He predicts a shift in 2026 where buyers leverage AI to control their journeys, transforming marketing dynamics.
AI Snips
Chapters
Transcript
Episode notes
Context Engineering Replaces Prompt Tricks
- Context engineering goes beyond prompt engineering by bundling instructions, data access, and tool permissions for agentic AIs.
- Effective AI tasks require defining the job, the needed data, and the tools the agent can use.
Design Tasks Before Automating With AI
- Treat AI like a human worker: define the task, give the right data, and provide necessary tools before automating.
- Prototype by having a person perform the task to discover missing data and tool needs prior to tooling the AI.
Give Models Just-Enough Context
- Limit context to what is relevant; too much data can time out or overwhelm the model.
- Use Goldilocks-sized context: not too little, not too much, to keep agentic processes performant.