Battlecard in Minutes: Drafting With AI, Then Adding the One Thing AI Can't
Objective: Given raw competitor intel bullet points, use an AI tool to draft a battlecard skeleton, then add the single human-sourced customer quote or proof point that the lesson says is what actually makes reps use it.
You're the PMM at Klaviyo and a competitor just shipped a pricing change. Reps need an updated battlecard before their next call, not next week.
Draft the battlecard skeleton with AI from the raw intel, then add the one real customer detail that makes it usable instead of generic.
Before you start
What you'll need
Free path (everything below is enough to finish)
Free tier handles a single-document draft in one prompt
Free, easy to link from a shared sales enablement folder
The process
1 step
Step 01 of 01
The lesson's example workflow is AI drafts the battlecard from competitive intel, then a PMM adds the one customer quote that killed that competitor in a real deal, that single human addition is what makes reps actually use it.
The raw intel says the competitor just cut its entry-tier price by 20%. What does the AI-drafted skeleton get right, and what's missing before a rep can use it on a call?
Procedure
- Feed Claude the raw intel: competitor name, the pricing change, one known feature gap.
- Prompt it to draft a battlecard skeleton: 'When they say X, you say Y' objection-response format.
- Review the draft for generic filler, same risk as any AI first pass.
- Add one real detail: an actual customer quote or deal outcome where this exact objection came up.
- Confirm the final battlecard has at least one line no competitor's AI tool could have generated, because it's specific to a real Klaviyo deal.
AI DRAFT SKELETON: 'When they say: [Competitor] is 20% cheaper on entry tier. You say: Klaviyo's entry tier includes SMS and advanced segmentation that [Competitor] charges extra for.' ADDED HUMAN DETAIL: 'Proof point: In the Nov 2025 Meridian Apparel deal, the prospect switched back to Klaviyo after 3 months on [Competitor]'s cheaper tier because their segmentation limits cost them an estimated $18K in missed Black Friday revenue.'
Healthy
The final battlecard has at least one line that references a real, dated deal outcome.
Unhealthy
Shipping the AI skeleton to reps with no edit pass, generic objection-handling language that could belong to any competitor.
What this means
AI collapses the drafting time from hours to minutes, the one thing it can't do is remember what actually happened in your last deal, that's the PMM's job.
So what do I do about it?
| Symptom | Action | Effort |
|---|---|---|
| Reps stop opening the battlecard after the first week | Check whether it has a real, specific proof point or just AI-generated objection templates | 30 min |
Final deliverable
A one-page battlecard skeleton drafted with AI, edited for generic language, with one real customer quote or deal outcome added by hand.
See a reference example
Coinbase battlecard update (excerpt) When they say: [Competitor] has zero trading fees. You say: Coinbase's spread is transparent and built into the quoted price, no surprise fee at settlement. Proof point: The Harrow Capital deal (Jan 2026) nearly went to [Competitor] until their finance team found $6K in hidden settlement spread over one quarter, they signed with Coinbase the same week.
Success criteria
You're done when you can:
- The AI-drafted skeleton follows a clear 'when they say / you say' objection-response format
- The final version includes at least one specific, dated, human-sourced proof point the AI could not have generated