Skip to content

Free GTM Skills / Discovery and diagnosis

GTM diagnostic

Produce the 30-day read of a B2B company's go‑to‑market: where it stands on product, market, customers, funnel, tracking, channels, assets and team, where the leaks are, and the quick wins to do first. Use when someone new takes over marketing or go‑to‑market, when a founder asks where are we at, what should we fix first, audit our marketing, we don't know what's working, do a current state assessment or what's the low hanging fruit, or at the start of any engagement before planning. Runs first, before the foundation skills; its findings tell you which to run.

By Matt Santos, BOM Growth. Free under the MIT license.

Install

npx skills add BOM-Growth/b2b-gtm-skills --skill gtm-diagnostic

Run it in your project folder, then ask Claude Code for gtm-diagnostic by name or in your own words.

skills/gtm-diagnostic/SKILL.mdView on GitHub

Goal: in two to three weeks, know enough to be dangerous. Write down what is true today, what is leaking, and the three to five things worth doing first. Then start doing them: the quick wins are also the diagnostic, because fixing the first leak shows you where the next one is.

This is a learning exercise, not a report. The output is one page the founder reads and agrees with, plus a task list.

Before you start

This skill works on its own. If the person has their own marketing brain folder, read all of it first; much of the context is already there. If they do not have one, ask for access to what exists: website, analytics, CRM, ad accounts, the last deck, the last three months of marketing output, and five to ten recent customer calls. Do not ask for a marketing brain and do not set one up unless they ask.

Ask the founder three questions on the first day and write the answers down verbatim:

  • What does winning look like in the next 12 months, in numbers?
  • What have you already tried, and what happened?
  • What do you think the problem is?

Tools this skill uses when the person has them. Read-only throughout; this skill changes nothing.

Area Tool What to look at
Customers and pipeline HubSpot, Pipedrive, or a spreadsheet Deal count by stage, source, close rate, cycle length, lost reasons, how many contacts have no next step
Calls Fathom, Granola The last five to ten customer and sales calls; what buyers say and ask
Website Google Analytics 4, Search Console, the site itself Traffic by source, top pages, form and booking conversion, what the homepage says in five seconds
Tracking Google Tag Manager, CRM source fields, UTMs Whether a lead can be traced back to a source at deal level, not just contact level
Paid Google Ads, Meta Ads, LinkedIn Campaign Manager Spend, cost per lead, search terms, what is on and off
Email and social Their email tool, LinkedIn page, Buffer or similar Cadence, open and click rates, what got engagement
Assets Google Drive or SharePoint Deck, one-pagers, case studies, brand files; what exists and what is current
Competitors Their websites, review sites, LinkedIn Who the company says it competes with and who it actually loses to

Steps

  1. Day 1 to 3: read everything. Product, website, deck, pricing, past plans, past reports. Do not form opinions yet. Keep a running list of questions.

  2. Day 3 to 7: listen. Watch or read the customer calls. Note the words buyers use for the problem, the objections, the competitors mentioned, and how the team describes the product. Talk to whoever sells and whoever delivers. If possible, talk to two customers.

  3. Day 5 to 10: pull the numbers. Fill in this table with what is available; write “unknown” where it is not, because the unknowns are findings too.

    Measure Last 90 days Source
    Website visits and top three sources
    Leads by source (form, booking, call, inbound email, outbound, event, referral)
    Leads that got a reply within one business day
    Meetings booked and meetings held
    Opportunities created, won, lost, and the top lost reason
    Average deal size and sales cycle length
    Paid spend and cost per lead by platform
    Content published and anything that got engagement

  4. Score the eight areas. One line each, 1 to 3, with the evidence. 3 means it works and is written down; 2 means it exists but leaks or depends on one person; 1 means missing.

    • Positioning and ICP: can a new hire say who it is for and why it is better?
    • Website: does the homepage say it, and does it convert?
    • Sales process: is there a defined path from first touch to close, with assets at each step?
    • CRM and tracking: is the pipeline real, and can a deal be traced to its source?
    • Lead follow-up: does every lead get worked, and how fast?
    • Channels: which ones produce, which are on by habit?
    • Content and proof: is there anything a prospect can read, and any proof they can check?
    • Team and tools: who does what, and what runs only in someone’s head?

  5. Find the leaks. A leak is a place where demand that already exists gets lost: inbound leads with no follow-up, forms that go to an inbox nobody reads, booked demos that do not show, deals that stall at one stage, paid spend on search terms that never convert, a homepage that does not say who it is for. List every one you can see, with a rough size (leads per month, dollars per month) and the fix.

  6. Pick the quick wins. Three to five things that take days, not months, and that the founder will notice: usually follow-up rules, a form fix, a homepage headline, turning off wasted spend, one piece of proof made public. Each with an owner and a date. Start them before the write-up is finished.

  7. Name the real gaps. Beyond the quick wins, which of the foundation and build playbooks does this company need, and in what order. Be direct about whether the gap is skill, team or channel; the fix is different for each.

  8. Write the one page. Headline finding in one sentence. Then: where it stands (the scores), what is leaking (the list with sizes), what we do first (quick wins with dates), what comes next (the playbooks in order), and the open questions. Frame everything as opportunity and next step, never as criticism of what was done before; the people who did it are in the room.

  9. Review it with the founder. Correct the facts, agree the quick wins, agree the order. Date it.

Rules

  • Read only. Change nothing in any system during the diagnostic, except the quick wins the founder has agreed.
  • Write down what is true, not what should be true. Unknowns are findings.
  • Numbers over impressions. Every score has a piece of evidence beside it.
  • Never criticise past work in the write-up. Say what the opportunity is and what to do next.
  • One page for the founder. Working notes can be as long as needed and stay separate.
  • Three weeks maximum. If it is taking longer, you are researching instead of deciding.

Output

Return the one-page diagnostic, the numbers table, the eight scores with evidence, the leak list with sizes, the quick-win task list with owners and dates, and the recommended playbook order.

If the person has their own marketing brain folder, save to it:

  • 5-current/priorities.md: the quick wins and the playbook order
  • 5-current/decisions-log.md: the date and the agreed first moves
  • 1-context/ files: any facts about the company, customers, offers or competitors you confirmed along the way
  • sources/notes/: the full working notes, dated

If they do not have one, return the diagnostic and the working notes as two files. Mention once that marketing-brain is the usual next step so nothing learned here gets lost; do not set one up unless they ask.

Next: the quick wins, then whichever of icp-and-wedge, positioning-doc, crm-setup or homepage-rewrite the diagnostic put first.

Find out if we’re a fit.

In a 30-minute discovery call, you’ll get honest, no-strings advice you can act on right away, and a clear picture of what working together would look like.

Book your discovery call