GTM Roles 14 min read

Claude Code for Sales Reps: The Complete Workflow Guide

6 practical Claude Code workflows for AEs and sales reps — account research, deal briefs, call review, outreach sequences, proposals, and win/loss analysis. No coding required.

Travis Hurst ·

TL;DR: Claude Code isn’t just for marketers and developers. If you’re an AE, BDR, or sales leader, there are at least 6 workflows that will change how you prep for calls, research accounts, write proposals, and review what’s working. This guide covers each one with the exact approach — no coding, no API setup, just practical playbooks you can run today. These are the same workflows I teach in my workshops to AEs who’ve never touched a command line.

In this article:

Why Sales Reps Need Different AI Workflows

Most “AI for sales” content falls into one of two traps.

Trap one: generic prompt lists. “Use ChatGPT to write cold emails!” Great — and those emails will sound like everyone else’s cold emails, because ChatGPT doesn’t know your company, your buyers, or your competitors.

Trap two: developer-oriented automation. “Set up n8n + Apollo + Claude API for an automated SDR pipeline!” Maybe useful if you have an engineer on your team. Not useful if you’re an AE trying to close deals.

There’s a gap between “paste a prompt into ChatGPT” and “build an automated sales infrastructure.” Claude Code fills that gap.

Here’s why: Claude Code reads local files. That means your competitor dossiers, your brand voice guide, your buyer personas, your call transcripts, your CRM exports — Claude Code can read them all simultaneously and produce output that’s informed by your actual context. Not generic. Not hallucinated. Grounded in your data.

And it runs locally. Your deal notes, your account research, your competitive intel — none of it leaves your machine unless you want it to.

I’ve taught these workflows to AEs from companies like Patch, Windfall, and ForceExperts. People who sell for a living, not people who code for a living. They walk away with working systems.

Here are the 6 workflows that matter most.

Workflow 1: Account Research Briefs

What it does: Takes a target account and produces a structured research brief — company overview, recent news, key stakeholders, pain points, personalization hooks, and a “why now” narrative for outreach.

Why it matters: Good AEs research before they call. But thorough research takes 15-20 minutes per account manually. Claude Code does it in a couple of minutes while you prep for something else.

What to tell Claude Code:

“Research [Company Name] for a sales call. Read their website — homepage, about page, product pages, careers page, and any recent press or blog posts. Build an account research brief that includes: company overview (size, funding, industry), their current positioning and messaging, key executives and their LinkedIn presence if findable, recent news or announcements, open job postings (especially roles that signal priorities), potential pain points based on what we solve, personalization hooks I can reference in outreach, and a ‘why now’ narrative explaining why this is a good time to reach out. Save in /accounts/.”

What makes this compound

The account research brief doesn’t disappear after one call. It becomes a reference file that feeds into deal briefs, proposal customization, and outreach personalization. Every workflow you run after this inherits the research you’ve already done.

And if you’ve already built buyer personas and competitor dossiers, Claude Code can cross-reference them — “this account matches our VP of Marketing persona, and they’re currently using [Competitor], so here’s how to position against them.”

Workflow 2: Multi-Source Deal Briefs

What it does: Combines call transcripts, email threads, Slack conversations, and CRM notes into one coherent brief — “here’s what’s actually happening in this deal.”

Why it matters: In complex B2B sales, deal context is scattered everywhere. Gong transcripts, email threads, Slack DMs with your SE, CRM notes you wrote three weeks ago. Before a deal review or QBR, you’re scrambling to piece it together. Claude Code synthesizes it all.

What to tell Claude Code:

“I need a deal brief for the [Company] opportunity. Here are the sources: [paste or point to call transcripts, email threads, CRM notes, Slack messages]. Synthesize everything into a deal brief that includes: deal timeline (key milestones and dates), stakeholder map (who’s involved, their role in the decision, their sentiment), current status and next steps, risk flags (anything that could stall or kill the deal), open questions we haven’t answered yet, and recommended actions for this week. Be specific — quote directly from the sources where relevant.”

What makes this compound

Your deal briefs build on your account research. If you ran Workflow 1 first, Claude Code already knows the company context and can weave it into the deal narrative. Over time, your deal folder becomes a living history of the relationship — not scattered across 5 tools.

Workflow 3: Call Review & Coaching

What it does: Takes call transcripts (from Gong, Chorus, or manual notes) and grades them against your team’s criteria — discovery quality, talk ratio, objection handling, next steps, and coaching recommendations.

Why it matters: Sales leaders can’t listen to every call. Most calls go unreviewed. With Claude Code, every call gets graded against the criteria that actually matter to your team.

What to tell Claude Code:

“Review this sales call transcript. Grade it on these criteria (1-5 scale): Discovery quality (did the rep uncover pain, timeline, budget, and decision process?), Talk ratio (what percentage did the rep talk vs. the prospect — target is 40/60), Objection handling (were objections addressed directly with specifics, or glossed over?), Next steps (clear commitment with date and action, or vague ‘let’s circle back’?), and Value articulation (did the rep connect our solution to the prospect’s specific pain, or give a generic pitch?). For each area, provide specific coaching notes with timestamps or quotes from the transcript. End with the top 2 things this rep should focus on in their next call.”

What makes this compound

The grading criteria become a custom skill — you build it once, and every call review follows the same standard. Over time, you can batch-analyze a quarter’s worth of transcripts to spot patterns: “Our team consistently scores low on discovery quality when selling to VP-level buyers. Here’s what’s happening…”

That’s where the win/loss analysis workflow picks up.

Workflow 4: Outreach Sequence Generation

What it does: Generates personalized cold outreach sequences (email + LinkedIn) based on your ICP, value props, and target persona — with A/B variations and personalization hooks.

Why it matters: Generic cold outreach gets ignored. But writing truly personalized sequences for every target account is time-intensive. Claude Code splits the difference — personalized output at scale, informed by your actual positioning and buyer data.

What to tell Claude Code:

“Generate a 5-email cold outreach sequence for [persona title] at [company type]. Use these inputs: our ICP definition in /personas/, our value propositions, and the account research brief for [specific company] in /accounts/. For each email: write a subject line with an A/B variation, keep the body under 125 words, reference something specific about the prospect’s company (from the research brief), connect one of our value props to a likely pain point for this persona, and end with a clear, low-friction CTA (not ‘book a demo’ — something easier). Also generate a LinkedIn connection request message and 2 follow-up messages that complement the email sequence without repeating it.”

What makes this compound

Your ICP definition, value props, and persona files are reusable across every sequence you generate. And the account research briefs from Workflow 1 feed directly into personalization. The more context files you’ve built, the more personalized every sequence becomes — without you re-explaining anything.

Workflow 5: Proposal Customization

What it does: Takes your standard proposal template, account research, and call notes — then generates a tailored proposal draft with customer-specific pain points, relevant case studies, and customized value propositions woven throughout.

Why it matters: Every AE has sent a proposal that was 80% boilerplate. The problem isn’t the template — it’s the time it takes to properly customize it. Claude Code does the customization in minutes, pulling from the context you’ve already built.

What to tell Claude Code:

“Customize our proposal for [Company]. Use the proposal template in /templates/, the account research brief in /accounts/[company]/, and these call notes: [paste or point to notes]. Tailor every section: rewrite the executive summary to reference their specific situation and pain points. In the problem statement, use language from the prospect’s own words (from the call notes). In the solution section, emphasize the capabilities most relevant to their needs. Match case studies to their industry and company size. Frame the ROI section using their specific metrics and goals. Keep the pricing section standard but adjust the ‘why this investment makes sense’ framing. Output as a clean markdown file I can convert to PDF.”

What makes this compound

Your proposal template, case study library, and account research all live as local files. Each proposal you generate builds on the research you’ve already done. And the proposals themselves become reference material — next time you sell to a similar company, Claude Code can reference what resonated before.

Workflow 6: Win/Loss Pattern Analysis

What it does: Synthesizes closed-lost (and closed-won) deals to identify patterns — why you’re losing to specific competitors, which objections aren’t being overcome, and what’s working in deals you win.

Why it matters: CRM “loss reason” dropdowns are useless. “Lost to competitor” tells you nothing. This workflow reads the actual deal notes, call transcripts, and email threads to surface the real patterns.

What to tell Claude Code:

“Analyze our closed-lost deals from last quarter. Here are the CRM exports and deal notes for [number] lost deals: [paste or point to files]. Identify patterns across these dimensions: which competitors we’re losing to and why (not just ‘lost to Competitor X’ — what specifically did they offer that we didn’t?), which objections came up repeatedly that our reps didn’t overcome, at which stage deals are stalling and what triggers the stall, which buyer persona or company profile we consistently lose, and any gaps in our positioning or product that the data suggests. For each pattern, recommend specific actions: new talk tracks, positioning changes, or product feedback. End with a ‘where we win’ analysis using our closed-won deals for comparison.”

What makes this compound

Win/loss analysis feeds directly back into your battle cards. If you discover you’re consistently losing to Competitor X on pricing, update the battle card’s objection handling section with new talk tracks. If you find that deals stall at the technical evaluation stage, update your discovery questions to surface that earlier.

The analysis also feeds into call coaching — once you know the patterns, you can grade calls specifically on whether reps are addressing the issues you’ve identified.

How These Workflows Compound

Here’s the part that makes this more than a list of 6 prompts.

Every workflow produces reusable files. Account research briefs. Deal histories. Call grading criteria. Outreach templates. Proposal customizations. Win/loss insights.

And every workflow can reference files that other workflows have produced.

Account Research → feeds into → Deal Briefs, Outreach, Proposals
Battle Cards     → feeds into → Call Coaching, Outreach, Proposals
Call Reviews     → feeds into → Win/Loss Analysis
Win/Loss         → feeds into → Battle Cards, Call Coaching

This is why the second month of using Claude Code for sales is dramatically better than the first. Your context library grows. Every file makes the next workflow faster, more personalized, and more accurate.

It’s the same compounding effect I describe in the battle cards workflow and the skills article — you’re building a system, not using a tool.

Getting Started

If you’re a sales rep who’s never used Claude Code, here’s where to start:

  1. Install Claude Code — the VS Code extension is the easiest path for non-technical users
  2. Create a project folder with subfolders: /accounts, /deals, /templates, /transcripts
  3. Run Workflow 1 first — pick your hottest prospect and build an account research brief. This gives you immediate value and builds a file that every other workflow can reference.
  4. Build your brand foundation — if your marketing team hasn’t already, create brand voice and competitor files. These make every sales workflow more accurate.
  5. Add one workflow per week — don’t try to run all 6 on day one. Start with account research, then deal briefs, then call review. Layer them as you get comfortable.

The whole thing runs locally. Your deal notes, your competitive intel, your client conversations — they stay on your machine.

FAQ

Do I need to be technical to use these workflows?

No. I’ve taught all of these to AEs who have never opened a terminal. Claude Code handles the technical parts — you provide the context (call notes, CRM data, account research) and review the output.

How long does each workflow take?

Your hands-on time per workflow is typically 2-5 minutes — paste a prompt, point to the files, review the output. Claude Code does the reading and synthesis. The account research brief is the fastest (paste a prompt, wait, review). The deal brief takes a bit longer because you’re gathering sources from multiple places first.

Can I use these with Gong/Chorus transcripts?

Yes. Export your call transcripts as text files and save them locally. Claude Code reads them like any other file. You can batch multiple transcripts for pattern analysis or review them one at a time for coaching notes.

What about CRM data — can Claude Code connect to Salesforce?

Claude Code can read CSV exports from any CRM. For live CRM connections, you’d use MCP (Model Context Protocol) — but you don’t need that to get started. Start with exports, and add live connections later when the value is proven.

How is this different from tools like Gong Insights or Clari?

Those platforms are purpose-built for specific workflows (call analytics, forecasting) and they’re good at what they do. Claude Code is more flexible — it can handle account research, deal synthesis, proposal customization, outreach generation, and call review in one tool. It’s also dramatically cheaper. The tradeoff: no automated data ingestion (yet). You’re the one gathering the inputs.

Can I share these workflows with my sales team?

Yes. The prompts, skills, and folder structure work for anyone on your team. If you put the project in a shared folder, everyone runs the same playbooks. This is how teams standardize their AI workflows — same quality bar, same output format, same institutional knowledge.

What if I already use ChatGPT for sales work?

You can keep using ChatGPT for quick, context-free tasks. Claude Code’s advantage is workflows that need local files — reading transcripts, referencing competitor dossiers, maintaining deal histories. The comparison article covers where each tool shines.

Build the System, Not Just the Workflow

These 6 workflows are the starting point — but the real value is the system they create together. Your account research feeds your proposals. Your call reviews feed your battle cards. Your win/loss analysis feeds your coaching. Every file makes the next workflow faster and more accurate.

This is one of the core tracks in my Claude Code workshops for GTM professionals. AEs, marketing leaders, and product marketers all work through these workflows with their own company data — and walk out with files they can use immediately. Small cohorts, real exercises, real deals.

If you want to keep learning at your own pace, the NativeGTM newsletter covers new sales and marketing workflows every week — from a practitioner who’s taught these to real teams, not a thought leader who just reads about them.

The sales reps who build these systems now are going to prep faster, personalize better, and close more. That’s not hype — it’s just what happens when your AI actually knows your business.

Want to build workflows like these?

The NativeGTM workshop is a hands-on, 2-day intensive where you build real AI workflows for your specific role.

See Workshops