A live competitive intelligence system has four parts: an AI-built battlecard per competitor, a diagnostic prompt for live deal situations, discovery questions that surface competitor weaknesses without attacking them, and a monthly 20-minute refresh that keeps it all current. Built once, it turns competitive knowledge from tribal memory into a tactical playbook any rep on the team can run.
Most enterprise AEs have a decent understanding of who their competitors are. A much smaller number can walk into a live competitive deal and immediately know what to do, what not to say, and how to shift the evaluation criteria. That gap is the system's job.
Part one: build the battlecard (20 minutes per competitor)
Gather live data first — recent news, review-site commentary, pricing pages — then structure it into a card built for deal moments rather than marketing summaries.
Prompt: "You are a competitive intelligence analyst. Build a competitive battlecard for [COMPETITOR] vs [MY COMPANY]. Structure: (1) a 30-second overview of what they are, (2) why they win deals, (3) why they lose deals, (4) the 3 questions I should ask in discovery to surface whether a prospect has experienced their weaknesses, (5) things I should never say when they are in the deal, (6) the one strategic play that consistently wins when they are the alternative."
Sections five and six are what separate a usable battlecard from a feature-comparison chart. Knowing what not to say — the claims that trigger defensiveness or invite a bake-off you lose — is worth more than another row of checkmarks.
Part two: diagnose competitive situations live
When a prospect mentions they are also evaluating a named competitor, the next conversation is the whole game. The wrong move is arguing features; the right move is shifting what the evaluation measures.
Prompt: "My prospect just mentioned they are also talking to [COMPETITOR]. Based on their known strengths and weaknesses, what is the most important thing I should do in my next conversation to shift the evaluation criteria toward where we win? What one question should I ask?"
Part three: discovery questions that expose gaps
The most effective competitive positioning never names the competitor. It asks discovery questions that lead the prospect to discover the weakness themselves. Ask AI to write three discovery questions per known competitor weakness, framed as genuine curiosity about the prospect's requirements — "how are you planning to handle [scenario the competitor handles poorly]?" — so the prospect reaches the conclusion without you appearing to trash anyone. Conclusions prospects reach themselves are the only ones that survive the competitor's counter-pitch.
Part four: keep it current
Competitive intel decays fast — pricing changes, features ship, positioning shifts. Set up alerts for each competitor's company name, leadership, and product names. Once a month, paste the accumulated alerts and recent review commentary into the model and update the sections that changed: why they win, why they lose, and any new landmines. Twenty minutes a month keeps every battlecard live; the alternative is discovering a competitor's new capability from a prospect, mid-deal, which is the most expensive possible way to learn it.
What to feed the system
Input quality decides battlecard quality. The best sources, in rough order of value: transcripts from deals where the competitor was actually in the evaluation, win-loss interview notes, recent review-site commentary from verified users, the competitor's own release notes and pricing pages, and analyst coverage. Marketing comparison pages — yours or theirs — are the weakest input, because they describe the deal both vendors wish they were in rather than the one the prospect is running. When a card feels thin, the fix is almost always another primary source, not a longer prompt.
Making it a team asset
Individual battlecards help one rep; shared ones compound across the team. Keep the cards in a common workspace, assign each competitor an owner for the monthly refresh, and add a standing agenda item where reps report what competitors actually said and did in live deals. Field intel from real evaluations is the highest-grade input the system will ever get, and most teams let it evaporate in Slack threads.
Frequently Asked Questions
How often should competitive battlecards be updated?
Monthly for active competitors, quarterly for occasional ones. The 20-minute alert-digest refresh catches most changes; a live-deal surprise should always trigger an immediate update regardless of schedule.
Should reps ever criticize a competitor directly in a deal?
No — direct criticism reliably backfires by making the prospect defensive of their shortlist. Discovery questions that surface the weakness as the prospect's own conclusion do the same work without the trust cost.
What belongs in a battlecard beyond features?
Why the competitor wins, why they lose, the questions that expose their gaps, the claims to avoid making, and the single play that wins head-to-head. Feature tables age fastest and persuade least.
Put It to Work
The battlecard, diagnosis, and discovery-question prompts above come from a library of 2,900+ B2B sales prompts across 16 categories, with a full competitive-positioning set inside Deal Strategy. Browse the library and build your first battlecard before your next competitive deal.