Workshop: competitor teardown
A cited, structured read on a competitor — then war-game the response with your audience.
Run a grounded competitor teardown
Load evidence, analyze it, generate the battlecard, and export sources.
Choose Files from the sidebar.
Good afternoon
Choose a surface from the sidebar to begin.
- 1Open FilesChoose Files from the sidebar.
- 2Add evidenceUpload the sample research source.
- 3Choose the intakeRoute it to grounded analysis rather than generic storage.
- 4Run analysisExtract themes and contradictions before writing the report.
- 5Generate the reportCreate a simulated PDF with source lineage.
- 6Export the evidenceDownload the analyst export and verify the source map.
Bring: a competitor's name. Time: ~45 min (part of it unattended).
Launch the audit (2 min)
Two speeds:
- Fast (minutes): Audit page → company name → structured teardown with cited sources.
- Deep (an hour, unattended): Programs → Deep Research → "Map [competitor]'s positioning, pricing, target segments, and recent moves. Cited sources only." It works while you do something else.
Read it like an operator (10 min)
When it's back, push past the summary:
What's their actual wedge — the thing they do that we structurally can't copy quickly? And what's ours against them?
Where is their pricing page making a promise their product likely can't keep? (Flag speculation as speculation.)
War-game it with the audience (15 min)
This is the move most teams never make — put the competitive question to the customer panel:
Build/select an audience of [the contested segment]. Focus group: "You're choosing between us and [competitor]. Walk through how you'd actually decide. What would each side have to say to win you?"
The decision criteria that come out of this group are your battlecard.
The battlecard (5 min)
Write the battlecard: their wedge, our wedge, the five decision criteria from the group in order of weight, our talk track for each, and the one claim of theirs we should publicly pressure.
Citations or it didn't happen
Everything in an audit/program about the real world carries a source link. If you see an uncited market claim, ask "source?" — and treat anything unsourceable as hypothesis, not fact. The synthetic panel tells you how customers decide; the web research tells you what's true out there. Keep the two straight and you'll never embarrass yourself in a client meeting.