GTM AI Podcast & Newsletter

GTM AI Podcast & Newsletter

The Playbook

Claude Code for Sales Teams

J Moss's avatar
J Moss
Mar 21, 2026
∙ Paid

Your VP of Sales walks into the QBR with a theory.

“We’re losing enterprise deals because procurement gets involved too late.” Everyone nods. It sounds right. It feels right. She’s been doing this for 15 years and her pattern recognition is sharp.

But it’s a theory. Not a finding. Nobody’s actually pulled the last 200 closed-won and closed-lost deals and looked at where procurement first appeared in the timeline, how that correlated with outcome, and whether it varies by deal size or industry. That analysis has been on the list for six months. It will stay on the list until someone with six free hours and strong SQL skills decides to prioritize it.

Claude Code doesn’t require either of those things. Feed it your data and describe what you want. Twenty minutes later, you have findings — not theories.


What You’ll Build

  1. A deal scoring model — built from your actual closed-won/lost history, not your CRM’s generic defaults

  2. A territory planning tool — turns your account list into a prioritized tier assignment in minutes

  3. A win/loss analysis report — surfaces patterns your team has been guessing at for quarters

  4. A commission calculator — lets reps see projected earnings in real time as deals move through stages

  5. An outbound sequence builder — personalizes multi-touch sequences by persona and account


Step 1: Understand What You’re Working With

Claude Code is Anthropic’s AI environment. You describe what you want to build — in plain English — and Claude figures out how to build it. No programming required. The skill is learning to be specific about two things: what goes in, and what you want to come out.

The 1 million token context window is the thing that makes this useful for sales teams. It means you can paste your entire closed deal history — two years of deals, every column from your CRM export — and Claude Code has it all in front of it at once. It’s not sampling. It’s reading the whole thing and finding patterns.

To get started: Go to claude.ai, start a new project, and open Claude Code. You need a paid plan (Pro or Teams).


Step 2: Build Your Deal Scoring Model

Export your CRM — every closed deal from the last 18–24 months, closed-won AND closed-lost. You want at minimum: company name, industry, employee count, annual revenue, deal size, sales cycle length, number of stakeholders touched, product/tier purchased, lead source, region, outcome (won/lost).

Save as CSV. Paste the full contents into Claude Code. Then:

I'm a VP of Sales and I want to build a deal scoring model from my historical pipeline data.
I've pasted our closed-won and closed-lost deals from the last 2 years below.

Please analyze this data and:
1. Identify which firmographic characteristics (company size, industry, revenue range) are
   most strongly correlated with wins vs. losses
2. Identify which behavioral signals (deal size, stakeholder count, lead source, sales cycle
   length) predict outcome
3. Build a scoring rubric with weighted variables — a simple point system I can use to score
   new opportunities at the top of my funnel
4. Flag any patterns that stand out as particularly strong predictors I might not have expected

Give me the rubric as a table: Variable | Value Range | Points. Keep the total score out of 100.

[PASTE YOUR CSV DATA HERE]

What you get back: A scoring rubric your CRM should have been running — something like “+15 points if company has 200–500 employees, +10 points if lead source is referral, -12 points if deal size is under $15K in a 60+ day cycle.” Variables weighted by actual correlation to your win rate.


Step 3: Your First Result

This is the moment that usually gets someone’s attention. The scoring rubric lands and it contradicts something the team believed. Maybe you’re losing mid-market deals disproportionately, but not for the reason anyone assumed. Maybe referrals don’t close faster — they close bigger. Maybe one industry vertical has a 22% win rate and you’ve never quantified that before.

Take the finding seriously. This is what the pipeline has been trying to tell you.


Keep reading with a 7-day free trial

Subscribe to GTM AI Podcast & Newsletter to keep reading this post and get 7 days of free access to the full post archives.

Already a paid subscriber? Sign in
© 2026 Coach K and J Moss · Privacy ∙ Terms ∙ Collection notice
Start your SubstackGet the app
Substack is the home for great culture