Agents now sit on both sides of the transaction. Martech Futurist | June 20, 2026
Get MarTech Futurist in your inbox. Subscribe to my once-weekly newsletter on LinkedIn.
AI agents now sit on both sides of the transaction. They rank the information that shapes a purchase, they buy on a person's behalf, and they sell inside answer engines and assistants. Forrester made the stakes plain for marketing leaders this week: brands market to machines as well as to people. Advantage goes to whoever the agent picks, and the agent decides before any human reads a message. Verification is the gate. A machine checks your product data and your claims in milliseconds, with no human in the loop, and what it finds determines whether your brand makes the shortlist.
Three themes run across this week's qualifying research:
Agents are becoming economic actors that read structured, verifiable data to decide what to surface and buy.
Brands now compete to be selected by agents at the first gate of discovery and purchase.
Real-time, unified data and journeys built as decision systems become the operating requirement.
Featured Insights
Forrester: AI Agents Are Your New Customer
Source: Forrester Blog | June 16, 2026
Chuck Gahun introduces business-to-agent (B2A) marketing. Agents skip the website and the narrative. They retrieve, validate, and surface structured information to answer engines and consumer agents, and they reward the sources that stay consistent across many interactions. Gahun cites Forrester's 2026 B2B Marketing Survey, in which only 24% of B2B marketing decision-makers plan to make their content visible and authoritative in AI-powered search and genAI tools. He sets three requirements: shape the agent's context across earned, owned, and paid sources; build information architecture clear enough that agents extract facts without inferring or hallucinating; and earn agentic loyalty by keeping product data consistent and verifiable across the web. Loyalty still rests on a relationship that compounds with trust. The party deciding whether to trust you is now a machine. Audit how your priority products render to an answer engine today. When an agent has to guess a spec or reconcile conflicting claims across your site and third-party sources, you are teaching it to favor a competitor.
Forrester: Google Goes All-In on an AI-Operated System
Source: Forrester Blog | June 18, 2026
Emily Collins reads Google's latest moves as a shift to an AI-operated system across its products, where intelligence runs the operation end to end. Three years after ChatGPT reset expectations for search, Google has regained position in the answer-engine race, even as OpenAI and Anthropic diversify the market. For marketers, the interface itself is dissolving into the model. The page recedes, an assistant assembles the answer, and the brand either appears in that answer or stays out of sight. Visibility becomes a function of how an AI system reads you, and that changes both the work and the measurement. Give answer-engine presence an owner and a budget, and judge it on its own terms. The analytics built for blue links will never tell you whether an AI answer included you.
CMSWire: Databricks Launches CustomerLake, an Agentic CDP
Source: CMSWire | June 16, 2026
At its Data + AI Summit, Databricks introduced CustomerLake, a customer data platform built on its lakehouse that unifies data, models, agents, identity resolution, audience building, and activation. The platform runs what Databricks calls infinity campaigns: continuous agentic loops that react to customer context in real time and replace the one-off campaign cycle. A data-and-AI infrastructure vendor has walked straight into martech. The signal underneath the launch carries further than the product. Orchestration is moving toward continuous, autonomous loops, and those loops run on unified data that agents can act on the moment context changes. Whoever owns the unified, real-time data layer holds the leverage in an agentic stack. Map where that sits in your own architecture before a renewal cycle settles it for you.
Forrester: Elevate Journeys Into Decision Systems
Source: Forrester Blog | June 18, 2026
Joana de Quintanilha argues that journey maps earn their value only when they connect to the decisions and investments they inform. Many teams sit on backlogs worth millions without a clear view of which journeys they improve and which they break. Her fix is to run journeys as decision systems that route data, set priorities, and tie experience changes to business outcomes. That same structure is what an agent needs to act on a journey in real time. The mapping exercise becomes a precondition for automation. Start with one high-value journey and attach two numbers to it: the revenue it influences and the cost of its current friction. A journey an agent can act on begins as a journey your own teams can measure.
Key Takeaways
B2A marketing belongs on the plan. Put one owner in charge of how your priority products read to an agent across earned, owned, and paid sources.
The discovery surface is now an AI system. Fund answer-engine visibility on its own, and judge it with metrics the blue-link analytics never captured.
Data readiness sets the ceiling. Continuous orchestration works on unified, real-time data, and stalls without it.
Journeys earn their keep when they carry numbers. Attach influenced revenue and friction cost to each priority journey before you automate a step.
None of this waits for a carefully planned roadmap. A brand that keeps its data clean and its claims honest is already doing the work an agent rewards, because the agent is checking exactly that. The advantage compounds quietly, in architecture and governance, long before a customer's agent ever runs the query.