How to Build Competitive Battle Cards with Claude Code (Step-by-Step)
A practitioner's guide to building professional competitive battle cards using Claude Code — no expensive tools required. Includes the exact workflow I teach in my workshops.
TL;DR: You don’t need Klue, Crayon, or any $15K/year competitive intelligence platform to build battle cards your sales team will actually use. With Claude Code and about 5-10 minutes of hands-on effort per competitor, you can go from zero to a production-ready battle card — complete with objection handling, discovery questions, landmines to plant, and even branded HTML output. This is the exact workflow I teach in my AI workshops for GTM professionals.
In this article:
- What You’re Actually Building — what separates a real battle card from a sales sheet
- The Workflow: 4 Phases — the full pipeline overview
- Phase 1: Competitor Research — building the dossier
- Phase 2: Your Brand Foundation — voice + personas
- Phase 3: Generate the Battle Card — the actual card
- Phase 4: Visual Output — branded HTML
- The Full Pipeline in One Session — how it compounds
- Why Claude Code Instead of Other Tools — honest comparison
- Getting Started — minimum setup
- FAQ
Most battle card advice falls into two camps.
Camp one: “Buy our tool.” Klue, Crayon, Playwise, SiftHub — they all want $15K-$30K/year to automate competitive intelligence. And they’re good at what they do. But if you’re a startup, a small team, or just a marketing leader who doesn’t have that budget yet? You’re stuck.
Camp two: “Here’s a template.” HubSpot, Gong, Kompyte — they give you frameworks (FIA! Talk tracks! Objection handling!) and a blank Google Doc. Useful in theory. But going from a blank template to a finished battle card that your sales team actually pulls up during calls? That’s a different problem.
There’s a third option nobody’s really talking about: build them yourself with Claude Code.
Not Claude.ai in a browser. Claude Code — the CLI tool that can scrape websites, read local files, and generate polished output in a single workflow. The difference matters, and I’ll explain why.
I’ve taught this workflow to 50+ GTM professionals across 3 workshop cohorts. Marketing leaders, AEs, product marketers, SEO directors — people who’ve never written a line of code. And they build their first real battle card in a single session.
Here’s how.
What You’re Actually Building
Before we get into the how — let’s be clear about the what.
A battle card isn’t a feature comparison table with green checkmarks. That’s a sales sheet.
A real battle card answers these questions for your sales team:
- “Why do we win against this competitor?” — not features, but positioning
- “What do they say, and how do I counter it?” — objection handling with actual talk tracks
- “What questions should I ask to expose their weaknesses?” — discovery questions
- “What seeds should I plant early in the deal?” — landmine questions
- “When should I let them have the deal?” — knowing when you don’t win
That last one is important. The best battle cards are honest. They tell your reps when the competitor is genuinely the better fit — because nothing kills credibility faster than pretending you win everywhere.
The Workflow: 4 Phases
Here’s the full pipeline. Each phase builds on the last, and Claude Code handles the heavy lifting.
PHASE 1: Competitor Research (scrape + dossier)
PHASE 2: Your Brand Foundation (voice + personas)
PHASE 3: Battle Card Generation (the actual card)
PHASE 4: Visual Output (branded HTML/PDF)Here’s the thing most people miss: Phases 1 and 2 produce reusable files. Your brand voice guide, your buyer personas, your competitor dossiers — they all live on your machine as local files. Every future workflow you run (not just battle cards) gets better because these files already exist. The first time through is the longest. After that, it compounds.
Phase 1: Competitor Research
This is where Claude Code earns its keep versus Claude.ai in a browser.
You’re going to point Claude Code at your competitor’s website and let it build a comprehensive dossier. Not a summary — a structured competitive intelligence document.
What to tell Claude Code:
“Research [Competitor] by reading their website — homepage, pricing page, about page, product pages, and any case studies or resources. Build a competitive dossier that includes: company overview (founding date, headcount, funding if available), their exact positioning statement and tagline, the specific differentiators they claim on their homepage, pricing tiers and model (or note if pricing is hidden), target customer segments they call out by name, their strengths relative to us, their weaknesses relative to us, and how they explicitly position against companies in our category. Save as a markdown file in /competitors/.”
What Claude Code does:
- Scrapes the competitor’s website (homepage, pricing page, about page, product pages)
- Reads their positioning and messaging
- Structures everything into a dossier format
- Saves it as a markdown file you can reference later
Your hands-on effort here is about 2 minutes — paste the prompt, hit enter, review the output when it’s done.
Why this matters
The dossier isn’t the battle card. It’s the intelligence that feeds the battle card. And here’s the compounding part — this dossier file doesn’t just serve battle cards. It becomes a reference for positioning work, sales prep, pitch decks, even blog content. Every file you build makes the next workflow smarter.
From my experience running this in workshops, the dossier phase takes a couple of minutes of your actual time — Claude Code does the reading while you grab coffee.
What a good dossier looks like
A solid competitor dossier covers:
| Section | What It Answers |
|---|---|
| Company Overview | Who are they? Founded when? How big? |
| Positioning | What’s their tagline? How do they describe themselves? |
| Key Differentiators | What do they claim makes them different? |
| Pricing | What do they charge? What model? |
| Strengths vs. You | Where are they genuinely better? |
| Weaknesses vs. You | Where do they fall short? |
| Head-to-Head | Side-by-side comparison on key dimensions |
| When They Win | What type of buyer chooses them? |
| When You Win | What type of buyer should choose you? |
Be honest in the weaknesses section. If the competitor is cheaper, say that. If they have a better free tier, say that. Your reps need the truth, not spin — they’re going to hear it from prospects anyway.
Phase 2: Your Brand Foundation
Here’s where most people skip ahead and it costs them. They go straight from competitor research to battle card. The result? Generic cards that sound like they were written by AI (because they were) and don’t reflect how your company actually talks.
You need two things before generating the battle card:
1. Brand Voice File
Tell Claude Code to analyze your own website and extract your brand voice — but be specific about what you want pulled.
“Read our website at [your-url]. Extract our brand voice in detail: the overall tone (formal vs. casual, technical vs. accessible), key themes we repeat across pages, power words and phrases we lean on, words and phrases we clearly avoid, headline patterns and structures we use, and how our CTAs are worded. Also pull the exact hex codes our brand uses, the specific fonts, take note of when each font is used (headlines vs. body vs. accents), and any other visual identity features that set this brand apart — logo placement patterns, icon style, image treatment, spacing conventions. Save everything as a brand voice and visual identity file in /brand/.”
This becomes a reusable reference file. Every battle card — and every piece of content — you generate going forward will sound and look like your company, not like generic AI output. This is one of those files that pays dividends across everything you do. Once your brand voice file exists, your custom Claude Code skills can reference it automatically, and every workflow from content creation to sales enablement gets the same consistent voice without you re-explaining it each time.
2. Buyer Personas
If you already have personas documented, point Claude Code at them. If you don’t, this is a good time to build them.
“Based on our website and positioning, create 2-3 buyer personas for our primary customers. For each persona include: their exact job title and seniority level, typical company size and industry, the top 3 pain points that would make them search for a solution like ours, their professional goals for the next 12 months, the specific criteria they’d use to evaluate vendors (price sensitivity, integration requirements, team size constraints, compliance needs), and the messaging hooks that would resonate most with each persona. Save in /personas/.”
Why both matter for battle cards
Your battle card needs to know who’s reading it (the persona) and how to sound (the brand voice). A battle card for an AE selling to a VP of Marketing reads differently than one for a BDR cold-calling a director of e-commerce.
And just like the dossier — these files compound. Your brand voice guide and personas feed into account research workflows, content creation, outreach personalization, and every other GTM workflow you build. You’re not just making a battle card right now. You’re building the foundation of a system.
Phase 3: Generate the Battle Card
Now you have three things in local files:
- Competitor dossier
- Brand voice guide
- Buyer personas
This is the context engineering part — and it’s why Claude Code is different from ChatGPT or Claude.ai for this workflow.
Claude Code reads all three files simultaneously and generates a battle card that’s informed by real competitive research, written in your brand’s voice, and targeted to your actual buyers. No copy-pasting between tabs. No re-explaining your brand every time. The context is already there.
What to tell Claude Code:
“Using the competitor dossier for [Competitor] in /competitors/, our brand voice guide in /brand/, and our buyer personas in /personas/ — create a competitive battle card for our sales team. Include these sections: Know Your Competitor (2-3 sentence overview with key stats), Quick-Glance Comparison (side-by-side table on 6-8 key dimensions), Why We Win (4 differentiators, each with a specific talk track an AE could say verbatim on a call), Where They Win (be honest — include what to say when a prospect brings up their genuine strengths), Discovery Questions (8-10 questions that naturally expose the competitor’s weaknesses without being obvious), Objection Handling (top 5 objections with counters that sound conversational, not corporate), Landmines to Plant (questions to ask early in the deal that create doubt about the competitor later), Ideal Prospect Profile (who we win with vs. who they win with), and a Bottom Line summary (3 sentences max). Write all talk tracks in a conversational tone — these need to sound like things a real person would say, not marketing copy. Save in /battlecards/.”
The sections that matter most
From running this in workshops, here’s what sales teams actually use during calls:
Objection handling — “They said Competitor X is cheaper. What do I say?” This is the section that gets pulled up mid-call. Make the counters specific and conversational, not corporate. Real talk tracks, not marketing copy.
Discovery questions — Questions designed to expose the competitor’s weaknesses without being obvious about it. “How important is [thing competitor is bad at] to you?” These need to feel natural, not scripted.
Landmines to plant — Seeds you drop early in the conversation that create doubt about the competitor later. “How much strategic support do you get from your current platform?” — a question that sounds innocent but sets up your white-glove advantage.
When to walk away — Genuinely. If the prospect is a $200/month budget and the competitor has a free tier? Let them go. Your reps will trust the battle card more if it’s honest about where you don’t win.
Phase 4: Visual Output
Markdown battle cards are useful for reference. But if you want something you can actually distribute to a sales team — something that looks professional and matches your brand — you need a visual version.
Claude Code can generate a single-file HTML battle card that uses your exact brand colors, fonts, and design system. And because you already built that brand voice and visual identity file in Phase 2, Claude Code already knows your hex codes, your font stack, and your design patterns. No re-explaining.
What to tell Claude Code:
“Create a polished HTML version of this battle card. Reference our brand visual identity file in /brand/ for exact hex codes, fonts, and styling conventions. Use our primary brand color for headers and section dividers, our secondary color for accent elements, and maintain our font hierarchy (headline font for section titles, body font for content). Include a subtle header with our company name, a clean table layout for the Quick-Glance Comparison, and collapsible sections for Objection Handling and Discovery Questions so it’s scannable. Output a single self-contained HTML file (all CSS inline) that can be opened in any browser or converted to PDF. Save in /battlecards/.”
The output is a branded, scannable, one-page-ish document your sales team can actually use. Not a 15-page PDF nobody reads. Not a Notion page buried in a workspace. A standalone file they can bookmark or print.
The Full Pipeline in One Session
Here’s the real power: once you’ve done this for one competitor, the second one is dramatically faster.
Your brand voice file and personas are already built. Those files are doing their job every time Claude Code references them. You just need a new dossier and a new battle card. The pipeline shrinks to:
Competitor 2: Paste prompt → Dossier → Battle Card → HTML (~5-10 min of your time)
Competitor 3: Paste prompt → Dossier → Battle Card → HTML (~5-10 min of your time)Each step is a prompt you paste, a button you hit, and a quick review of the output. Claude Code does the heavy lifting while you focus on whether the positioning and talk tracks actually reflect reality.
I’ve watched workshop participants go from zero to three complete, branded battle cards in a single session. And these aren’t people who code for a living — they’re marketing leaders, AEs, and product marketers.
And that’s just battle cards. The brand files, competitor dossiers, and personas you built along the way? Those become the foundation for content workflows, account research, sales prep, and a dozen other GTM playbooks. Every file makes the next workflow faster, more accurate, and more on-brand. That’s what I mean by compounding — the system gets better the more you use it.
Why Claude Code Instead of [Other Tool]
Fair question. Here’s the honest breakdown:
vs. Klue/Crayon ($15K-$30K/year): Those platforms are better if you need ongoing automated monitoring — tracking competitor website changes, aggregating review sites, pulling news. If you’re a 500-person company with a dedicated competitive intel team, they make sense. If you’re a startup or mid-market team that needs solid battle cards without the contract? Claude Code gets you 80% of the way there.
vs. ChatGPT: ChatGPT can write battle cards. But it can’t read your local files, can’t scrape competitor websites in a structured way, and can’t generate HTML output that matches your brand system — all in one workflow. You end up copy-pasting between tabs. Claude Code runs the full pipeline from a single interface. For a deeper breakdown on where each tool shines, I wrote a full comparison for marketing work.
vs. Claude.ai in the browser: Similar AI, different capability. Claude.ai can build a battle card from information you paste in. Claude Code reads files from your machine, scrapes websites, generates file output, and chains the whole workflow together. It’s the difference between a conversation and a production tool.
vs. a blank template: Templates give you structure. Claude Code gives you structure and content and brand compliance and visual output. The template is one step in the process. Claude Code handles all of them.
What This Looks Like in Practice
In my workshops, we use a real company as the working example. Not a hypothetical — a real brand, real competitors, real market.
Participants walk in with their own company context. By the end of the session, they’ve built:
- A competitor dossier for their #1 competitor
- A brand voice file for their own company
- A complete battle card with talk tracks, objection handling, and discovery questions
- An HTML-rendered version they can share with their sales team
The feedback that keeps coming back: it’s not the battle card itself that’s valuable — it’s the system. The dossier, the brand voice file, the personas — those files stick around. They feed into every workflow you build after this. When you go to write a positioning doc or prep for a sales call, Claude Code already knows your brand, your competitors, and your buyers. You’re not starting from scratch every time. That’s the compounding effect — and once people experience it, they don’t go back to one-off prompting.
Getting Started
If you want to try this yourself, here’s the minimum setup:
- Install Claude Code — the VS Code extension is the easiest starting point if you’re non-technical
- Create a project folder with subfolders for
/brand,/competitors,/personas, and/battlecards - Start with Phase 1 — pick your biggest competitor and build the dossier
- Build your brand foundation in Phase 2 — this is the step that pays the biggest dividends, because these files power everything else you’ll build
- Generate your first card and iterate from there
The whole thing runs locally on your machine. No data leaves your computer unless you want it to. Your competitive intelligence stays yours.
FAQ
How long does it take to build one battle card?
Your actual hands-on time is about 5-10 minutes per competitor — paste a prompt, hit enter, review the output, move to the next step. Claude Code does the reading and writing. The first competitor takes a bit longer because you’re also building your brand voice file and personas, but those are one-time investments that make every future workflow faster.
Do I need to know how to code?
No. That’s the whole point. I’ve taught this to marketing leaders, account executives, and product marketers who have never written a line of code. Claude Code handles the technical parts — you provide the strategy and business knowledge.
How do I keep battle cards updated?
Re-run Phase 1 (competitor research) quarterly, or whenever a competitor makes a big move. Your brand voice and personas rarely change, so updates are fast. The dossier files stay on your machine — just tell Claude Code to refresh them. Because the files are already structured from the first run, updates take a fraction of the original effort.
Can I use this for more than just sales battle cards?
The workflow pattern is the same for any competitive content: competitive positioning for pitch decks, win/loss analysis frameworks, competitor comparison pages for your website, or even internal strategy documents. And the files you build here — dossiers, brand voice, personas — feed directly into other workflows like content creation, account research, and sales enablement. Battle cards are the most common starting point because the output is immediately useful, but the real value is the system you’re building underneath.
What if I have 10+ competitors?
Start with your top 3. Build those battle cards first, refine the process, then expand. Most sales teams only need battle cards for the competitors they actually face in deals — which is usually 3-5, not 20.
How is this different from what Anthropic’s official guide shows?
Anthropic’s guide for Claude.ai focuses on connecting CRM data (HubSpot) and using web search to generate battle cards in the browser. It’s a good approach if you want a quick card. The workflow I teach goes deeper: building reusable competitive research files, extracting your actual brand voice and visual identity, creating buyer personas, and generating branded visual output — all running locally through Claude Code. The difference is a one-off card vs. a repeatable competitive intelligence system that gets better every time you use it.
What does a complete battle card include?
The format I teach has 9 sections: Know Your Competitor (overview), Quick-Glance Comparison (table), Why We Win (4 differentiators with talk tracks), Where They Win (with counters), Discovery Questions, Objection Handling, Landmines to Plant, Ideal Prospect Profile, and a Bottom Line summary. Each section is designed for a specific moment in the sales process.
Build the System, Not Just the Card
Battle cards are the most tangible output of this workflow — your sales team uses them immediately. But the real unlock is what happens after your first battle card is done.
You now have a brand voice file that keeps every piece of content on-brand. Competitor dossiers that feed into positioning, pitches, and strategy docs. Buyer personas that sharpen everything from cold outreach to product messaging. These files become the foundation of your AI-powered GTM system — and every new workflow you add builds on what’s already there.
This is one of 20+ playbooks I teach in my Claude Code workshops for GTM professionals. Battle cards, account research, brand voice extraction, content pipelines, sales prep — each one builds on the last. Small cohorts, real company exercises, no slides-and-lecture format. You leave with working files, not just notes.
If you want to keep learning at your own pace, the NativeGTM newsletter drops new workflows, prompts, and use cases every week — practical stuff you can run the same day you read it.
The tools are here. The playbooks exist. The question is whether you’re going to keep doing this stuff manually or start building the system.
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