The title shows up in job ads, agency pitches and tool marketing, and each uses it a little differently. Below: what the role builds, how fast it is growing according to independent job-ad data, which skills it needs, and how the work is moving from no-code builders into AI coding tools connected over MCP.
What is a GTM engineer?
A GTM engineer turns a go-to-market strategy into running systems. A sales rep works the accounts and a RevOps manager owns the CRM and reporting. The GTM engineer builds the machinery between them: where target accounts come from, how they are enriched and scored, which signals trigger an action, and how the result lands in a sequence or with the right rep.
The role sits between operations and engineering. It needs enough sales context to know what a good lead looks like, and enough technical skill to connect APIs, write a query or a script, and keep a pipeline running when a data source changes.
Where did the role come from, and how fast is it growing?
The term was popularised by Clay, which credits itself with coining it around 2023. Independent job-ad data shows fast growth from a small base. Kyle Poyar's Growth Unhinged analysis, using Sumble data, counted about 45 GTM engineer job ads a month in summer 2025, against fewer than ten a month in 2024. The same analysis found roughly one GTM engineer for every 92 SDRs, and about 45% of the roles at agencies rather than in-house teams.
Bloomberry's review of 1,000 GTM engineering job ads found postings up 205% year on year, a median advertised salary of $127,500, and SQL and Python both required in 38% of ads. These are job-ad counts, not headcount, and the base is still small. The direction is the same in both sources.
What does a GTM engineer build in a week?
Mostly pipelines: chains of steps that start with a raw input and end with a person contacted for a reason. A typical one has five steps.
| Step | What happens | Typical method |
|---|---|---|
| Signal | An event at a target account is detected: a job ad, a new executive, a funding round | Signal feeds, registry data, job portals |
| Enrichment | Company and contact data is added: size, technologies, open roles, verified email | Data providers, waterfall enrichment, custom API calls |
| Classification | A model or a set of rules decides whether the account fits, and why | A prompt with criteria, SQL filters |
| Sequence | Qualified contacts enter a sequence whose first line comes from the signal | A sending platform |
| CRM | Replies, tasks and results reach the rep who owns the account | CRM and chat integrations, webhooks |
Around the pipelines sits maintenance: fixing a data source that broke, tightening a classification prompt that lets bad fits through, and measuring which signals turn into meetings.
Which skills and tools does the role use?
The skill set mixes data work, light engineering and sales judgement.
- Data: SQL, large spreadsheets, and a working sense of data quality and decay.
- Code: Python or JavaScript for scripts and API calls.
- APIs and webhooks: connecting tools that were not built to talk to each other.
- Classification prompts: writing criteria a model can apply consistently, then testing them on a hand-checked sample.
- Sales context: the ideal customer profile, the buying group, and what a rep needs in order to act.
- Orchestration: workflow tools such as n8n, and increasingly AI coding environments such as Claude Code, Cursor and Codex.
How does a GTM engineer work from Claude Code, Cursor or Codex?
Through MCP. The Model Context Protocol, introduced by Anthropic in 2024 as an open standard, lets an AI tool call external systems as tools. A GTM engineer connects data, enrichment and sending platforms as MCP servers, then describes the task to the agent in the IDE: find companies matching a profile, enrich them, classify them, and push the qualified ones into a sequence.
That changes how platforms get used. Leadspicker's knowledge base puts it plainly: use exactly the slice you need, the data, the enrichment or the sending, and orchestrate it inside a larger pipeline you build yourself, so that the platform is one step in the workflow rather than the workflow. The pipeline lives in code the engineer controls and versions, and each platform contributes the step it does best.
The trade-off is ownership. A pipeline you write is a pipeline you maintain. A team without someone to own it is usually better served by the platform's built-in flows.
Leadspicker connects over MCP to Codex, Cursor, Claude Code and other MCP-capable tools, so a GTM engineer can use one slice of the platform as a step in a pipeline they build themselves.
Frequently asked questions
What does a GTM engineer build in a typical week?
Pipelines that take a signal or a target list, enrich and classify the accounts, push qualified contacts into sequences and route results to the CRM. The rest of the week goes to maintenance: fixing data sources, tuning classification prompts and measuring which signals turn into meetings.
Which skills does a GTM engineer need?
SQL and a scripting language such as Python, API and webhook integration, prompt writing for classification, and enough sales context to recognise a good lead. Bloomberry found SQL and Python both required in 38% of the GTM engineering job ads it analysed.
How does a GTM engineer use an MCP server?
The engineer connects platforms such as a lead database or a sending tool to an AI coding environment over MCP. The agent can then call their functions as tools, so one request or script can search, enrich and launch a sequence without switching apps.
- Kyle Poyar, Growth Unhinged: Do you need a GTM engineer? (2025)
- Bloomberry, I analyzed 1,000 GTM engineering jobs (2025)
- Clay, GTM engineering: what it is (2026), cited only for the origin of the term
- Anthropic, Introducing the Model Context Protocol (2024)
- GTM playbook for 2026 by Leadspicker, chapters: Using Leadspicker from your AI tools (MCP); Pricing and plans; Custom data sources
