GTM Wire Glossary

Short definitions for GTM operators reading AI product news.

Agent

An AI system that can take a goal, use tools, and complete a multi-step task. For GTM teams, an agent is useful only when the inputs, permissions, review point, and success metric are clear.

Agentic Workflow

A workflow where AI does more than suggest text. It gathers context, chooses a next step, uses a system, or prepares work for approval. Common GTM examples include account research, campaign QA, lead routing, support triage, and offer preparation.

Connector

A permissioned link between an AI tool and a business system such as CRM, email, files, calendar, ads, support, or analytics. The GTM question is which fields the AI can read, which fields it can write, and who approves changes.

Context

The information an AI system can use while producing an answer or taking an action. In GTM work, useful context usually means account data, product docs, campaign history, support history, meeting notes, and current pipeline state.

Grounding

The practice of tying an AI answer to known source material instead of letting it answer from general model memory. GTM teams need grounding when outputs affect claims, pricing, routing, segmentation, or customer communication.

Human Review Boundary

The line where AI assistance stops and a named person approves the output. This matters most for customer-facing claims, CRM writes, budget changes, legal language, qualification decisions, and account ownership.

Model

The AI system that produces the answer or action. Model choice matters, but GTM teams should focus first on workflow fit, data access, review points, cost, latency, and measurable output quality.

Orchestration

The coordination layer that moves work across tools, teams, and approval steps. In GTM, orchestration often connects intake, enrichment, drafting, routing, review, publishing, and measurement.

Retrieval

The process of finding relevant business information before an AI system answers. Retrieval quality determines whether the AI sees the right policy, account record, product note, campaign result, or support thread.

Tool Use

When an AI system calls another system to search, calculate, write, update, send, create, or retrieve something. Tool use is where GTM risk increases, because the AI may affect live systems or customer-facing work.

Workflow Metric

The number that proves the AI-assisted workflow improved. Useful GTM metrics include accepted outputs, rework rate, cycle time, conversion rate, meeting quality, routing accuracy, deflection rate, and pipeline impact.