Competitive narrative war games: how buyers hear you in AI answers

Buyers do not read your positioning doc. They ask an assistant for a shortlist and a rationale. A narrative war gamepressure-tests how that compression behaves under stress: pricing questions, security objections, regional nuance, and “who is the safe choice” framing.

Setup: prompt sets, not vibes

Build twenty to forty prompts that mirror real discovery: evaluation criteria, compare X vs Y, migration risk, total cost, “best for mid-market,” and objection prompts your sales team logs weekly. Tag each prompt by funnel stage and persona.

Adjudication rules

• Fact vs inference: assistants blend both; your team must separate what is verifiable on your site from what is model generalization.
• Harm threshold: define what counts as a loss—wrong SKU, wrong region, wrong compliance posture—before you debate tone.
• Competitor fairness: document where competitors are accurately described vs where your own materials are thin.

Three rounds that surface strategy

Round 1 — Cold read

Run the prompt set silently; each participant writes: top risk, top opportunity, one surprise. No groupthink yet.

Round 2 — Red team as competitor

Half the room argues the competitor story using only public sources. The goal is not cynicism—it is to find the holes in your proof stack.

Round 3 — Commitments

End with owners: web fixes, content, analyst outreach, partner proof, pricing clarity. Each commitment has a measurable definition of done before the next diagnostic.

Outputs leadership can reuse

• One-page “assistant narrative” summary: how the category is framed.
• Risk register tied to public claims and owners.
• Competitive gap map: where you will not out-copy—you will out-prove.