Become a GTM EngineerPremium

Outcome: You can describe the GTM data stack end to end and name the owner of every hop.

1. 1.1What a GTM engineer actually doesThe role sits between revenue and engineering: you build the systems that find, qualify and reach buyers.12 min
2. 1.2The GTM stack, hop by hopSource → clean → enrich → research → score → route → send → measure, and what breaks at each hop.14 min
3. 1.3Metrics a GTM engineer is judged onCoverage, match rate, cost per qualified row, reply rate and pipeline per 1,000 rows.10 min

Outcome: You can build a clean, deduped account + contact table from three messy sources.

1. 2.1Sourcing accounts and peopleWhere good lists actually come from, and how to avoid inheriting someone else's junk.13 min
2. 2.2Identity resolution and dedupeOne row per real company, one row per real human — the hardest unglamorous problem in GTM.15 min
3. 2.3Normalisation recipes you will reuse foreverNames, titles, seniority, geography and headcount bands, all cleaned with plain formulas.12 min

Outcome: You can build a cost-guarded enrichment table that runs unattended.

1. 3.1The Clay mental model (Find, Enrich, Transform)Tables are spreadsheets with a job queue attached. Columns are functions. Rows are entities.11 min
2. 3.2Advanced Clay: Web Scraping, Gen AI, & HTTP APIHow to step outside pre-built native integrations using the HTTP API, Webhook triggers, and Gen AI classifiers.16 min

Outcome: You can write prompts that return strict JSON with citations and safe fallbacks.

1. 4.1Prompts that return structured, usable outputJSON schemas, hard rules, refusal paths and why 'be creative' is the enemy.16 min
2. 4.2Grounding, citations and verificationMake the model quote the page, then check the quote before you send anything.13 min
3. 4.3Building your own prompt libraryVersioning, evaluation sets, and knowing when a prompt actually got better.12 min

Outcome: You can score an account list and generate first lines that survive a human read.

1. 5.1Signals: what to watch and how fresh it must beFit, intent and timing signals — with decay windows so you stop sending stale news.13 min
2. 5.2Scoring and routingA score a rep can argue with, and routing rules that never double-assign an account.14 min
3. 5.3Personalisation that survives a human readVariable design, first-line generation, and the QA gate before anything sends.15 min

Outcome: You can ship a production pipeline with monitoring, and prove its impact.

1. 6.1Orchestration with n8nWebhooks, queues, retries and rate limits — turning a table into a running system.15 min
2. 6.2CRM sync without corrupting the CRMField mapping, upsert rules, source stamping and the fields you must never overwrite.13 min
3. 6.3QA, monitoring and proving your impactThe pre-send checklist, the daily health query, and how to package the work as a portfolio.14 min

Outcome: You can stand up a sending infrastructure that lands and diagnose it when it doesn't.

1. 7.1Sending infrastructure from zeroDomains, mailboxes, DNS records and the topology that keeps your main domain safe.16 min
2. 7.2List hygiene as a deliverability leverVerification tiers, catch-all handling, suppression logic and the bounce budget.13 min
3. 7.3Diagnosing a channel that stopped workingThe decision tree for 'replies died last week', with the tests that isolate each cause.12 min

Outcome: You can build a sub-5-minute inbound loop and a PQL model on product events.

1. 8.1Speed to lead: the sub-5-minute loopForm capture, instant enrichment, qualification and routing without a human in the path.15 min
2. 8.2Website and third-party intentDe-anonymisation, page-weighted scoring, and turning visits into a defensible outbound trigger.13 min
3. 8.3Product-qualified leads and lifecycle signalsEvent schemas, PQL thresholds, expansion triggers and churn-risk plays built on usage data.14 min

Outcome: You can enrich from any API that has docs, without waiting for a native integration.

1. 9.1HTTP API columns: enrich from anythingCalling any documented API from Clay — auth, headers, dynamic bodies and response parsing.16 min
2. 9.2Pagination, rate limits and big jobsCursor loops, batching, backoff, and how to move 100k rows without getting banned or billed twice.14 min
3. 9.3Building your own enrichment serviceWhen to stop buying data and ship a small internal endpoint — with caching, cost control and a contract.15 min

Outcome: You can package the skill into a role or an offer and run it like a system.

1. 10.1Positioning and the portfolio that gets repliesHow to describe the work, what to show, and the three artefacts hiring managers actually read.12 min
2. 10.2The interview loop and the take-homeWhat each stage tests, the live-build trap, and how to answer the cost question well.13 min
3. 10.3Freelance pricing and the weekly operating cadencePackaging offers, scoping retainers, and the rhythm that keeps a GTM system healthy.13 min

Outcome: You can hit the same coverage on a fraction of the credits, and prove the saving.

1. 11.1How credits actually get burnedThe cost model behind every column, the four ways teams waste credits, and the unit metric to track.14 min
2. 11.2The low-credit tactic listTwenty concrete tricks — free-first ladders, gating patterns, sampling, caching and manual escape hatches.18 min
3. 11.3Budgets, alerts and the monthly cost reviewAllocating credits per play, catching a burn spike within the hour, and the review that keeps spend flat.13 min