Why this is hard to get right
Imagine you're a sales enablement manager at a fast-growing SaaS company. You've just learned that 40% of your losses last quarter went to the same two competitors. Your VP of Sales pulls you aside after the QBR: "We need battle cards by end of week."
You open ChatGPT and type: "Write a battle card comparing us to [Competitor]." What comes back looks polished — but it's useless. It lists feature comparisons anyone could find on a G2 page. It doesn't mention the specific objection your reps hear every single time: "We're already standardizing on their platform." It doesn't include the three questions your best rep uses to expose the competitor's weaknesses in enterprise accounts. It reads like a press release, not a weapon.
This is the fundamental problem with vague competitive prompts. The AI doesn't know your product's real edge. It doesn't know your buyers are VP-level RevOps leaders who care about implementation speed above all else. It doesn't know your reps have between 30 and 90 seconds to scan a resource before jumping on a call.
So you spend two hours editing. You loop in the product team. You schedule a call with your top AE to get the real talk tracks. What should have taken 20 minutes takes two days — and the output still doesn't feel quite right.
The frustration is real because the stakes are high. A bad battle card doesn't just fail to help — it can actively hurt, giving reps false confidence or the wrong messaging in a high-stakes evaluation. Sales teams often abandon battle cards entirely because the ones they've seen haven't been worth the paper they're printed on.
A well-structured prompt changes this entirely. When you give the AI the right inputs — competitor, buyer profile, deal stage, specific objections, rep skill level, desired format — it produces a draft that's 80% ready in 5 minutes. You spend your time refining and validating, not starting from scratch.
Common mistakes to avoid
Naming the Competitor Without Naming Your Product
If the AI doesn't know what you sell, it can only describe the competitor — not why you win. Always include your product category, your key differentiator, and the specific buyer problem you solve better. Without this, the output becomes a one-sided competitor profile.
Skipping the Rep's Perspective
Battle cards are tools for human beings in a specific moment of stress. Forgetting to specify the rep's experience level, time constraints, and how they'll use the card produces overly long or overly simplistic output. A card meant to be scanned in 60 seconds before a call must be built with that constraint baked in.
Using Marketing Language in the Prompt
Phrases like 'industry-leading' or 'best-in-class' in your prompt contaminate the output with the same vague language. Use plain, direct language that describes what your product actually does and what outcome buyers get. The AI will mirror your specificity.
Omitting Real Objection Phrasing
Asking for 'common objections' produces textbook responses. Pasting in the actual words a prospect said last week — 'I don't want to rip out what we have' — produces talk tracks that feel authentic and rehearsed, not robotic.
Forgetting the Disqualification Criteria
The most credible battle cards tell reps when NOT to compete. Omitting this section signals to the AI that you want pure cheerleading content, which reps distrust. Including a 'when to walk away' section produces a more balanced, trustworthy card that gets used more often.
The transformation
Write a battle card comparing us to Salesforce so our sales team can beat them in deals.
**Act as a senior sales enablement strategist.** Create a one-page competitive battle card for a B2B SaaS revenue intelligence platform competing against Salesforce Revenue Cloud.
**Target user:** Mid-market account executives (2-5 years experience) closing deals with VP of Sales and RevOps buyers at 200-1,000 employee companies.
**Structure the battle card as follows:**
1. **One-line positioning statement** (our key differentiator in plain language)
2. **3 situations where we win** (specific deal scenarios with trigger phrases)
3. **Top 3 objections and talk tracks** ("Salesforce is our system of record" / "We already have forecasting" / "Switching costs are too high")
4. **3 landmines to plant** (questions reps can ask to expose Salesforce limitations)
5. **When NOT to compete** (deal profiles where Salesforce is the right fit and we should walk away)
**Tone:** Direct, confident, conversational — written for a rep to scan in 60 seconds before a call. Avoid marketing language.Why this works
Role Framing
Opening with 'Act as a senior sales enablement strategist' activates a specific set of domain conventions in the AI's output — structured sections, rep-centric language, and strategic framing — rather than the default tone of a marketing writer or general analyst.
Buyer Specificity
Naming the exact buyer role (VP of Sales, RevOps) and company size (200-1,000 employees) gives the AI concrete context to calibrate vocabulary, pain points, and decision-making dynamics — producing talk tracks that feel like they were written for a real person, not a job title.
Verbatim Anchoring
Including actual objection phrases in quotation marks forces the AI to engage with real language rather than hypothetical scenarios. This produces responses that match what reps hear in live conversations, making the talk tracks immediately usable.
Format Constraint
Specifying a one-page, five-section structure with a 60-second scan requirement overrides the AI's tendency to pad output with context paragraphs and summaries. The constraint produces a card that is dense, scannable, and actionable.
Strategic Nuance via Disqualification
The 'when NOT to compete' instruction signals to the AI that credibility and strategic judgment matter more than cheerleading. This single instruction shifts the entire tone of the output from promotional to trusted-advisor — which is exactly how great sales enablement content should read.
The framework behind the prompt
Competitive battle cards are a core artifact of the sales enablement discipline, which sits at the intersection of product marketing, revenue operations, and frontline sales coaching. The best battle cards apply a framework borrowed from military intelligence: they don't just describe the enemy — they anticipate the enemy's moves and give your team a playbook to counter them.
The most effective battle cards draw on two proven sales frameworks. The first is Challenger Sale methodology, which emphasizes teaching buyers something they don't know rather than pitching features. A good battle card doesn't just tell reps "we're better at X" — it gives them questions and insights that reframe the buyer's thinking before the competitor ever comes up.
The second framework is SPIN Selling, which structures sales conversations around Situation, Problem, Implication, and Need-Payoff questions. The "landmine questions" section of a strong battle card is SPIN in action — it gives reps the Implication and Problem questions that expose the competitor's weaknesses without requiring the rep to attack directly.
Research in sales psychology consistently shows that reps who feel confident and prepared in competitive situations close at significantly higher rates than those who improvise. Battle cards reduce cognitive load at the worst possible moment — when a prospect throws a curveball — by giving reps a pre-loaded response they can access without thinking.
The challenge is that most battle cards are built by marketers for marketers, not by salespeople for salespeople. The best ones are co-created with frontline reps using real objections, real win/loss data, and real deal context — which is exactly what a well-structured AI prompt can help systematize.
Prompt variations
Act as a sales development trainer. Create a one-page competitive cheat sheet for SDRs who encounter [Competitor] during cold outreach to Director-level buyers at 50-500 person B2B companies.
Include:
- One-sentence differentiator (jargon-free, for a cold call)
- 3 discovery questions that uncover dissatisfaction with the competitor without name-dropping
- 2 quick responses to 'We already use [Competitor]' that pivot to a meeting request
- Green light signals — 3 things a prospect says that mean they're open to switching
Tone: Energetic, simple, and confidence-building. Designed for someone with under 12 months of sales experience.
Act as an enterprise sales strategist. Build a competitive battle card for a 6-12 month enterprise deal cycle where [Competitor] is the incumbent vendor. Buyers include IT Security, Procurement, and VP-level business owners.
Structure:
- Stakeholder-by-stakeholder positioning — the unique value message and top objection for each buyer persona
- Procurement-specific talk track for total cost of ownership conversations
- IT Security objection handler for data residency and compliance concerns
- Champion enablement section — 3 points your internal champion can use to build consensus internally
- Competitive displacement checklist — 5 questions that expose migration risk from the incumbent
Format: Detailed but scannable. Designed for an AE to share selectively with internal champions.
Act as a product marketing expert. Create a feature-level competitive battle card focused specifically on [your feature category, e.g., 'reporting and analytics'] versus [Competitor].
Target audience: Sales engineers and technical AEs fielding detailed product questions in evaluation stages.
Include:
- Side-by-side capability matrix (3-5 key capabilities, plain-language descriptions)
- 3 demo talking points that highlight your advantage without triggering a spec-sheet debate
- Responses to 'But [Competitor] has X feature' — 2 bridging statements per capability gap
- Technical landmine questions to ask in discovery that expose [Competitor]'s architectural limitations
Tone: Precise, technically credible, and measured. Avoid superlatives.
When to use this prompt
Sales Enablement Managers
Build a full library of battle cards covering your top 3-5 competitors, each tailored to specific buyer personas and deal stages, without scheduling weeks of cross-functional workshops.
Competitive Intelligence Analysts
Rapidly convert raw competitive research — win/loss data, G2 reviews, analyst reports — into structured, rep-ready content that sales teams can act on immediately.
Product Marketers
Translate product positioning documents and feature comparisons into field-facing language that resonates with AEs and SEs who need talk tracks, not marketing decks.
Sales Managers Prepping for a Big Deal
Generate a targeted, deal-specific battle card 24 hours before a competitive evaluation, incorporating the specific objections and stakeholders your rep has already encountered.
Revenue Operations Leaders
Standardize how reps handle competitive situations across the team, reducing reliance on tribal knowledge and ensuring consistent messaging in every deal cycle.
Pro tips
- 1
Specify the rep experience level in your prompt — a battle card for a new SDR needs simpler language and more prescriptive talk tracks than one for a senior enterprise AE.
- 2
Include 1-2 real objections your reps have actually heard verbatim — this forces the AI to write talk tracks that match real conversations instead of imagined ones.
- 3
Name the specific buyer role and their top business priority, not just the company size — a CFO evaluating total cost of ownership needs different talking points than a VP of Sales focused on adoption.
- 4
Add a 'freshness constraint' to your prompt by specifying a product version or recent differentiator — this prevents the AI from leaning on outdated or generic product comparisons.
A single battle card is useful. A structured library is a competitive advantage.
Step 1: Rank your competitors by deal frequency. Pull your last 90 days of closed-lost data. Which competitors appear most often? Build cards for the top 3-5 first — don't boil the ocean.
Step 2: Segment by deal type, not just competitor. You may need different cards for the same competitor depending on deal size, buyer persona, or product module. A card for a 10-seat SMB deal against HubSpot looks nothing like one for a 500-seat enterprise deal.
Step 3: Run a talk track validation session. After generating your first draft, schedule a 30-minute session with your top 2-3 reps. Ask: 'Is this the real objection you hear, or a sanitized version?' Their edits transform a good card into a great one.
Step 4: Embed battle cards in your CRM. The best battle card is the one your reps can find in 10 seconds. Add competitor-specific cards as pinned notes on opportunity records, or trigger them automatically when a competitor field is populated. Adoption lives or dies on accessibility.
Step 5: Track which cards get used — and which don't. Low usage is signal, not failure. It means either the format isn't working, the content isn't trusted, or reps don't know it exists. Review usage monthly and iterate.
The fastest way to upgrade a battle card from generic to authoritative is to inject real win/loss data into your prompt.
What to include:
- The top 3 reasons you won deals against this competitor (from closed-won analysis)
- The top 3 reasons you lost (from closed-lost calls or CRM data)
- Verbatim prospect quotes from loss interviews, if available
- The deal profiles where your win rate is highest (company size, industry, use case)
Example prompt addition:
"Incorporate the following win/loss context: We win most often when the buyer has outgrown spreadsheets and tried the competitor's self-serve tier. We lose most often when the competitor is already embedded with the IT team and our champion is a business user without budget authority. A lost prospect told us: 'Their implementation team was already onsite and switching felt too risky.'"
This single paragraph of context shifts the entire output. The AI now writes talk tracks that address the real decision dynamics — not hypothetical ones.
Where to get this data if you don't have formal win/loss interviews: Ask your top 3 AEs to each submit 5 bullet points from recent competitive deals. Even rough qualitative data produces dramatically better output than asking the AI to guess.
Use this checklist to evaluate any battle card — AI-generated or hand-written.
Structure
- [ ] Fits on one page or is scannable in under 90 seconds
- [ ] Has a clear, jargon-free positioning statement
- [ ] Covers at least 3 specific deal scenarios (not generic situations)
- [ ] Includes objection handlers written in the prospect's language, not yours
Strategic Depth
- [ ] Contains at least 2-3 discovery questions reps can use to expose competitor weaknesses
- [ ] Addresses the competitor's most compelling strength honestly (not dismissively)
- [ ] Includes a disqualification section — when to walk away or de-prioritize
Rep Usability
- [ ] Written at the vocabulary level of the target rep, not the product team
- [ ] Free of marketing copy and superlatives ('best-in-class', 'industry-leading')
- [ ] Includes at least one 'if they say X, you say Y' script that feels natural to speak aloud
- [ ] Has a clear 'last updated' date so reps know how fresh the information is
If your battle card misses 3 or more of these, it's worth regenerating with a more specific prompt. Reps can tell the difference between a card built with real context and one that was rushed.
When not to use this prompt
Don't use this prompt pattern when you're in the earliest stage of competitive research and don't yet know your own product's differentiated value. If you can't articulate why you win, the AI will fill that gap with generic claims that may actively mislead your reps.
This prompt is also less effective for emerging competitors with limited public information. In that case, first use a separate research prompt to gather and synthesize everything available, then feed that summary into this battle card prompt as explicit context.
Finally, avoid using a single battle card prompt for multiple competitors at once. Each competitor deserves a focused, dedicated prompt — blended output is always weaker than targeted output.
Troubleshooting
The AI output reads like a marketing brochure, not a sales tool
Add this instruction to your prompt: 'Do not use marketing language, superlatives, or adjectives. Write every sentence as if a rep is speaking it aloud on a call.' Then specify the rep's experience level and the exact moment they'll use the card — before a call, during an objection, in an email. Situational context changes the register of the entire output.
The objection handlers feel scripted and unnatural
Replace generic objection labels with verbatim quotes from actual prospect conversations. Instead of 'handle the 'we already have a solution' objection,' write 'handle this exact objection: 'We standardized on Salesforce last year and my team would push back hard on anything new.' The more specific the input, the more conversational and usable the response.
The battle card is too long and reps won't read it
Add a hard constraint to your prompt: 'The entire output must be under 300 words and use bullet points only — no full paragraphs.' You can also ask for a 'summary version' first (one sentence per section) and then ask the AI to expand only the sections that need more detail. Length is a prompt problem, not an AI problem.
How to measure success
A successful battle card output should pass the "60-second rep test": an experienced AE should be able to scan it before a call and immediately identify 2-3 things they'll actually say.
Look for these quality signals: objection handlers written in first-person conversational language (not third-person summaries), discovery questions that are specific enough to ask verbatim, a positioning statement under 20 words, and a "when not to compete" section that demonstrates honest strategic judgment.
If your reps read the card and say "this is exactly what I hear" — rather than "this is technically correct" — the prompt worked.
Now try it on something of your own
Reading about the framework is one thing. Watching it sharpen your own prompt is another — takes 90 seconds, no signup.
a rep-ready competitive battle card
Try one of these
Frequently asked questions
Revisit battle cards every 90 days at minimum, or immediately after a major competitor product launch, pricing change, or when your win/loss data shows a new pattern. A stale battle card is often worse than none — it gives reps false confidence with outdated information.
Yes, but you'll get better results if you first ask the AI to summarize the competitor's known positioning, then use that summary as input context in your battle card prompt. This two-step approach gives the AI consistent information to work from rather than making assumptions.
Add one sentence specifying your industry and the buyer's top regulatory or operational concern — for example, 'Buyers in healthcare are primarily concerned with HIPAA compliance and implementation risk.' This single addition shifts the entire tone and content of the output to match the industry context.
Include pricing context if your reps encounter it in deals — for example, 'Our list price is typically 20% lower but we lose on perception of enterprise readiness.' Giving the AI the real pricing dynamic produces more honest, useful talk tracks than asking it to guess the competitive pricing landscape.
Add an explicit word count constraint to your prompt — '250 words maximum' works well for a one-page card. You can also instruct the AI to use bullet points only, no full paragraphs, which naturally compresses the output without losing the key information.