So you want to be a GTM engineer.

The tools worth using, the ones to skip, and how the work actually gets done. Written by people doing the job every day, not selling a course.

A GTM engineer is a go-to-market person who can build, not an engineer sent to help sales. It's the same shift DevOps went through: infrastructure became code and a new role appeared. Revenue is going through it now. With AI and a small set of good tools, one person can do the work of a team, faster and with better data.

Scroll for more

How the work gets done

One loop, five steps, run every week. The tools below only matter if you work this way.

1

Spot the signal

Job changes, funding rounds, tech installs, hiring. 20 high-intent accounts a day beats a list of 10,000.

2

Research with AI

Before any touch, your stack already knows what they do, what tools they run and what probably hurts.

3

Reach out

WhatsApp, LinkedIn DMs, a call with real context. Not another templated email into the void.

4

Test claims fast

10 to 20 page and ad variants a week. The nos tell you which claim works. Hook rate first, then CTR, then CPA.

5

Automate what works

Version your prompts, build the missing tools, wire signal to outreach. Only automate what's proven.

Show only:

The stack

You don't need all of these. Start with the ones marked start here, add the rest when you hit the bottleneck they solve.

Orchestration and data hub

The centre of the stack. Everything else feeds into it or gets triggered from it.

ClayFree tierStart here

The hub most GTM engineers build around. A programmable spreadsheet connected to 50+ data providers for enrichment, scoring and routing. Treat it like a database with an orchestration layer, not a CRM plugin.

n8nFree

Visual workflow automation and the glue between signals, enrichment, scoring and outreach. Open source and self-hostable, so there's no per-task pricing when your volume grows.

Managed alternative to n8n with a huge integration library. Lower ceiling but faster to start. A good first automation tool if you don't want to host anything.

The original connector. Simpler than n8n or Make, more limited, and the pricing stings at volume. Fine for quick single-trigger automations. Most people outgrow it.

Before you buy anything, try a Google Sheet with Apps Script. You can call APIs, scrape data and build lookup functions with no dependencies and no monthly bill. It covers more than you'd think.

AI and LLM layer

You need at least one model API. This is what writes your research briefs, scores accounts and drafts your first lines.

Claude APIPaidStart here

Anthropic's models. Strong at long-context research, structured data extraction and writing that doesn't read like a robot. Use it for account research, ICP scoring and personalised outreach.

The other default. Function calling for structured outputs, fine-tuning for domain tasks, vision for parsing screenshots and documents. Most off-the-shelf integrations support it first.

Search built into the model. Feed it a company name, get back current research with citations. Removes the separate search-then-summarise step in automated account research.

Very fast inference for open models. Use it when you need answers in milliseconds, like live scoring or enrichment inside high-volume pipelines.

Code and build tools

The engineer part of the job. These let you ship your own dashboards, scrapers and internal tools without waiting on a product team.

CursorFree tierStart here

AI code editor. Describe what you want, get working code. Build dashboards, scrapers and custom integrations without being a full-time developer. This is the tool that makes the role possible.

WindsurfFree tier

Same idea as Cursor. Try both, keep whichever clicks. The habit of building your own tools matters more than the editor you pick.

Anthropic's coding agent in your terminal. Give it a task and it plans, writes and runs code across your whole project. The best option for scaffolding a complete tool from one description.

RetoolFree tier

Drag-and-drop internal tools. Connect databases, APIs and spreadsheets, get dashboards and admin panels without writing frontend code. Good for partner portals.

ReplitFree tier

Browser IDE with an AI agent. No local setup, deploys in a click. Handy for quick prototypes, webhook endpoints and lightweight tools.

Describe a UI, get a deployable React component. A fast way to spin up landing pages, dashboards and forms for claim testing.

Enrichment and data providers

Turning a domain or a name into a full record: contact details, company data, tech stack.

Apollo.ioFree tier

Contact database with 200M+ contacts and basic sequencing. The free tier is generous. Most people use it as a data source piped into Clay rather than on its own.

Hunter.ioFree tier

Email finder and verifier. Give it a domain, get verified addresses back. Simple, reliable, and the API is easy to build on.

Company and contact enrichment: revenue, headcount, tech stack, industry. Was the gold standard, now bundled into HubSpot. The API still works on its own.

GDPR-compliant email enrichment. Finds and verifies professional emails algorithmically instead of from a stored database. Matters if you sell in the EU.

Raw people and company data over API, 1.5B+ person profiles. For building your own enrichment pipeline when Clay's built-in providers don't go deep enough.

LushaFree tier

Direct dials and email addresses, with a Chrome extension for LinkedIn prospecting. The free credits are enough to test whether it's worth paying for in your market.

Snov.ioFree tier

Email finding, verification and drip campaigns in one, cheaper than Apollo. The technology search is useful for technographic prospecting.

Accurate email finder with a bulk export tool for Sales Navigator searches. Good accuracy for the price.

Scraping and data extraction

When the data you need isn't in any provider's database, you go and get it yourself.

ApifyFree tier

Scraping as a service, with pre-built scrapers for Google Maps, LinkedIn, TripAdvisor and hundreds more. Rent the scraper before you build one.

LinkedIn automation and scraping. Pulls profiles from Sales Navigator searches, auto-connects, auto-messages. LinkedIn's rate limits are real, so go easy. Pairs well with Clay.

Chains LinkedIn scraping, email finding and data cleaning into one pipeline. More robust than PhantomBuster for multi-step workflows.

BardeenFree tier

Browser automation from a Chrome extension. Scrape a site, push the data to a sheet or CRM, trigger workflows. The low-code option for simple scraping jobs.

OutscraperFree tier

Google Maps extraction: reviews, popular times, contact details, competitor data. Feed it Place IDs, get structured data back. Useful for local market research.

FirecrawlFree tier

Turns any website into clean, LLM-ready markdown. Built for feeding competitor pages, job boards and news sites into an AI research pipeline.

Backends and databases

Your tools need somewhere to keep data. These are cheap or free and take minutes to set up.

SupabaseFree tierStart here

Postgres with auth, storage and realtime built in. The backend for custom dashboards, partner portals and internal tools. The free tier covers a lot.

AirtableFree tier

Spreadsheet-database hybrid. Structure without SQL. Limited at scale, but quick for tracking pipeline and campaign logs.

NeonFree tier

Serverless Postgres that branches like Git and scales to zero. A real database with nothing to manage. Good Supabase alternative if you don't need the extras.

With Apps Script attached, Sheets is a free API-connected database. For early pipeline tracking, enrichment logs and quick dashboards it is genuinely enough.

CRM and pipeline

You still need a system of record. Pick one that stays out of your way.

AttioFree tier

Modern CRM with a flexible data model, auto-sync from email and calendar, and a proper API. Built by people who got fed up with Salesforce and HubSpot.

FolkFree tier

Lightweight relationship CRM with a Chrome extension that pulls contacts from anywhere. Good for founders who don't want enterprise software.

HubSpotFree tier

The free tier is genuinely useful: contacts, deals, email tracking, basic reporting. It gets expensive fast beyond that. Free HubSpot with Clay as the brain is a solid combination.

Built for outbound. Calling, email and SMS inside the product, and the best pipeline view for a small team doing volume. Feels made by salespeople, not enterprise architects.

Landing pages and creative

Claim testing needs pages and ads out the door in hours, not weeks.

FramerFree tier

Design-to-live website builder. Publish fast, iterate faster. The right tool for shipping a pile of landing page variants in one night.

WebflowFree tier

More powerful than Framer, steeper learning curve. Pick it when your pages need a CMS or dynamic content.

CarrdFree tier

One-page sites, €19 a year for pro. Headline, subhead, CTA, live in five minutes. The cheapest way to put a claim in front of real traffic.

Free video editing. Phone-shot, UGC-style ads edited in CapCut beat polished brand creative in paid social, consistently.

AI video avatars. Personalised video at scale without recording yourself 500 times. Feels odd, but the results are good enough to test.

Voice cloning and generation. Voicemail drops, voice notes and audio content at scale. The quality is now hard to tell from a real recording.

Analytics and tracking

If you're testing claims, you need to see what people actually do on the page.

PostHogFree tier

Product analytics, session replay, feature flags and A/B testing in one, with a generous free tier. Open source and self-hostable.

HotjarFree tier

Heatmaps and session recordings. Shows where people click, scroll and give up, which tells you why a page isn't converting.

Server-side conversion tracking for Meta ads. Without it your attribution is guesswork. Track leads, purchases and custom events. Worth the setup time.

Lightweight, privacy-friendly analytics with no cookie banner. Enough for landing page tests when you don't need PostHog's depth.

Signal sources

Signals beat lists. This is where the timing comes from.

Pulls buying signals from community activity, product usage, social and job changes into one feed you can act on.

Still the richest B2B signal source: saved searches, lead recommendations, job change alerts. Automate it with PhantomBuster or Captain Data.

Technographics. When a target installs a competitor's tool, you know, and you show up with context. BuiltWith for deep history, Wappalyzer for live detection.

Free signal capture for anything the paid tools miss: funding announcements, hiring posts, product launches. Pipe the alerts into your automation layer.

LinkedIn monitoring. Job changes, posts on chosen topics, engagement with competitors, growth milestones. Triggers your outreach at the right moment.

Tells you which companies visit your site, what they read and how often, before anyone fills in a form.

B2B intent data at scale. Tracks which companies are researching specific topics across the web. Expensive, but strong for finding accounts already in a buying cycle.

Shows which companies are comparing you and your competitors on G2. Someone reading category reviews is about as warm as intent gets.

Sales intelligence from the Kaggle founders. Tech stacks, projects and org charts across 2.6M+ companies, with signals pulled from job boards, filings and social. Clean data, updated daily.

Outreach and distribution

Where the pipeline actually gets made. Infrastructure first, then volume.

Cold email infrastructure

Deliverability is infrastructure. Get this wrong and nothing else in your outbound matters.

The default for volume cold email. Unlimited sending accounts, automated warmup, smart rotation and a deliverability dashboard.

Instantly's main competitor, with better multi-channel features, a unified inbox and API access. More setup, more flexibility at scale.

Cold email with personalised images and video, and liquid syntax for deep personalisation. A middle ground between volume and quality.

Bulk domain and mailbox setup. Dozens of sending domains and accounts in minutes. The plumbing behind high-volume cold email. Pair it with Instantly or Smartlead.

Multi-channel sequencing

Email on its own is noisy. These coordinate LinkedIn, email and other channels in one sequence.

LinkedIn, email and X in coordinated sequences, with if-then branching based on engagement. European company, takes GDPR seriously.

LinkedIn automation at scale. Rotate across multiple accounts, coordinate campaigns, track engagement.

Cloud-based LinkedIn automation with smart sequences, warmup and CRM integration. Safer than browser extensions.

For the right audience, WhatsApp beats email by a wide margin. Template messages, interactive flows, real conversations. Not for every market, but where it fits, it wins.

Ads and paid acquisition

Paid channels are the fastest way to test claims and capture demand that already exists.

Run 10 to 20 creative variations a week. Kill anything that misses your CPA target after two to three times target spend. UGC-style creative wins.

Capture existing demand on search. Best for high-intent keywords where people are already looking for a solution.

Expensive per click, but unmatched B2B targeting by job title, company size, industry and seniority. Only makes sense for high-ACV products.

AI voice agents for outbound calls. Use them for qualification and booking, not closing, and listen to the recordings.

What to avoid

The pattern behind all of these: “AI SDR” products that are just faster spam, monolithic outbound tools that feel like progress but don't convert, and six SaaS seats where a Google Sheet with Apps Script would do the job.

Spray and pray at scale

More emails does not mean more pipeline. Signal-driven outbound at lower volume converts 5 to 10 times better.

Tools before method

Spend two weeks doing GTM by hand first. Build pages, write outreach, make calls, track what converts. Then buy tools for the bottlenecks you actually hit.

Copying playbooks blindly

What worked for a Series B company with 50 reps won't work for you. Build your own playbook from what converts in your market.

Over-automating too early

Do it manually until you understand what works. Automate the thing that's proven, not the thing you hope will work. Premature automation scales mistakes.

Ignoring deliverability

Your best email means nothing in spam. Warm your domains, set up SPF, DKIM and DMARC, watch your sender reputation. Infrastructure before creativity.

Data without action

A Clay table with 10,000 enriched rows is worthless if nobody messages them. Build the action into the pipeline, not after it.

Your first month

Four weeks from nothing to a working pipeline machine. Every step uses free tiers or small budgets.

Week 1: Foundation

Sign up for Clay on the free tier. Build one enrichment table: 50 target accounts, enriched with company data, tech stack and recent news. Write a Google Apps Script that pulls company info into a spreadsheet. Send 10 personalised messages using what you found. Track every response.

Week 2: Signals

Set up three signal sources: Google Alerts for your ICP keywords, LinkedIn Sales Navigator saved searches, and one technographic trigger like BuiltWith or Wappalyzer. Connect them to a central sheet or Clay table. Start building your daily feed of high-intent accounts.

Week 3: Velocity

Build five landing pages with different claims for the same ICP, using Framer, Carrd or v0. Run traffic to all of them, even at €5 a day each. Build an n8n or Make workflow that connects signal to enrichment to notification. You now have a pipeline machine, not a pipeline process.

Week 4: Compound

Write three prompts of your own: account research, personalised first lines, ICP scoring. Version them and A/B test the outputs. Set up Instantly or Smartlead for email infrastructure. Connect everything: signals feed Clay, Clay feeds AI enrichment, AI feeds outreach. You're now a GTM engineer.

Build your own tools

Tools you can build yourself in 15 to 60 minutes with Claude Code. Open one, copy the prompt, paste it into your terminal.

These are starting points, not scripts. Change them, combine them, or use them as a spark for something else entirely. The best GTM tools are the ones built for your exact workflow.