The traditional B2B go-to-market model, consisting of siloed sales and marketing teams, manual lead routing, and gut-feel forecasting, was already showing its age. AI is accelerating the reckoning.

Earlier this year, Priya Gil, CMO at customer engagement and marketing platform provider Iterable, created a role at her company that didn’t have a name until recently: the GTM engineer. “A GTM engineer orchestrates AI agents across marketing and sales,” Gil explained. “They own the knowledge base, the systems layer, how agents interact with each other, and the entire lead flow process,” Gil says.

The role isn’t about writing code. It’s about understanding the technical infrastructure, building purpose-specific agents, and ensuring those agents connect across functions to complete end-to-end workflows without constant human intervention. The catalyst, Gil says, was straightforward. Expecting every designer or content marketer to become an AI expert isn’t realistic. Having one person who bridges strategy and systems and then enables teams to build agents customized to their workflows is.

That framing aligns with what Mary Shea, a former Forrester analyst and founder of New York–based Meerkat, a provider of AI ‘teammates’ that listen in on business conversations and mines them for follow-up tasks and relationship insights, predicted years before the term ‘GTM engineer’ existed. “These teams don’t manage humans; they manage agents,” Shea says. “Enablement teams are transforming into this function as well, educating AI agents to become team members rather than just training salespeople,” Shea says.

From Silos to Systems

The structural implications extend well beyond a single new job title.

Tim McCormick, a revenue operations executive, points to RevOps as the connective tissue making this shift possible. Five years ago, sales ops managed Salesforce, marketing ops managed HubSpot, and the two functions rarely coordinated. That fragmentation is unsustainable when AI orchestration requires integrated data and synchronized workflows across marketing, sales, and customer retention.

“RevOps teams are essentially becoming the engineering layer for go-to-market,” McCormick says. In the setup he describes, an inbound lead triggers a sequence of automated decisions: enrichment agents pull firmographic data, qualification agents score against the ideal customer profile, routing agents assign to the right rep, and outreach agents draft personalized communications based on the prospect’s digital footprint. Human involvement comes at the review stage, not the processing stage.

The hiring profile is shifting in parallel. Chris Tilton, a cybersecurity marketing veteran, says he stopped recruiting traditional business development representatives who follow call scripts. Instead, he’s seeking what he calls technical BDRs (business development representatives): people with technical degrees or backgrounds who can blend product knowledge, customer success instincts, and sales judgment. The reasoning is straightforward: buyers arrive 70% of the way through their decision process, already armed with AI-assisted research. They want a technical conversation immediately, not a scripted qualification call.

Where the Efficiency Case Is Sharpest

Based upon Shea’s observations, in leading organizations, AI-generated forecasting is further pushing manual forecasting to extinction. Buyer engagement signals, such as event attendance, content downloads, email response time, and sentiment, produce probabilistic forecasts that don’t depend on sales rep optimism. “Forecasting becomes off the table for individual contributors and managers,” Shea says.

McCormick also sees sales cycles compressing from nine months to seven months on average through systematized workflows. Real-time data sync, automated account-based marketing, and aggregated intent signals eliminate the gaps where deals historically stalled between human handoffs.

The Human Reconfiguration

None of this means headcount disappears. It means headcount reconfigures.

Gil predicts that growth in team size slows considerably, while the profile of the people hired changes sharply. “Instead of five generalist marketers, you might have two strategists with deep expertise who manage a team of agents,” she says. The people who thrive in that model need technical fluency, comfort with AI coaching, and the judgment to recognize when AI outputs are wrong or sloppy.

Shea is predicting a structural shift in the sales team’s composition that’s been in place for years. The traditional ratio has been roughly 7 account executives to 1 sales engineer. Shea believes that’s a reversal of the ratio toward six sales engineers to one closing account executive. That’s because buyers arrive educated and want technical depth, not pitches. Sales engineers can provide that at a lower total compensation than senior account executives.

However, the organizations that are getting this wrong appear to be doing so badly. “Companies that mandate ‘cut your team in half because AI can do it’ are failing spectacularly.” She’s observed CMOs forced to lay off content teams, only to rehire months later when AI-generated output underperformed. The organizations succeeding are investing in education, helping existing teams learn to work with AI as a collaborator rather than treating headcount reduction as the primary objective.

Tilton adds an important nuance that matters especially in B2B markets where relationships carry weight. CISOs, he noted, can identify AI-generated outreach immediately. What cuts through is technical substance delivered by people who understand the buyer’s specific environment. AI accelerates that preparation; he described reducing enterprise sales research that once took weeks to 10 minutes using AI analysis of S-1 filings and earnings reports, but it doesn’t replace the judgment that makes the conversation credible. “That’s not replacing human judgment; it’s amplifying it,” Tilton says.

What This Means for Revenue Leaders

McCormick’s counsel to leaders navigating this transition: build or invest in RevOps first. Without that unified function spanning marketing through customer success, the orchestration layer that makes AI agents effective at scale has nowhere to operate. “Don’t treat this as a tool implementation,” he says. “Treat it as an organizational redesign.”

Gil’s parallel advice centers on the GTM engineer role itself, whatever an organization chooses to call it. The function of bridging business strategy and technical infrastructure, she argues, becomes the force multiplier that determines whether AI agents deliver measurable efficiency or generate activity. Organizations that diffuse that responsibility across existing roles risk ending up with neither.

The broader picture Shea paints is one where the dividing line between winning and losing revenue organizations isn’t which AI tools they purchase. It’s whether they’ve deliberately built the human-AI hybrid model, with genuine specialists managing agents the way experienced managers develop junior employees, and with clear governance over what the agents are authorized to do, what they’re required to escalate, and how their outputs are validated.

That validation is essential. AI agents hallucinate. They produce plausible-sounding outputs that don’t hold up against the actual buyer context. The organizations building durable GTM infrastructure aren’t simply deploying agents and measuring throughput. They’re building oversight alongside that automation because the two are inseparable if the goal is to improve revenue performance.