The teardown is competitive intelligence's core artifact: one document that renders a rival legible - who they are, where they are going, how they sell, where they are exposed, and what you should do about it. Most teardowns fail their readers in one of two ways: they are screenshots with adjectives (evidence without synthesis), or confident narrative with no sources (synthesis without evidence). This article is the anatomy of one that works, section by section - the structure the Market Intelligence agent produces on demand, and the standard worth holding any teardown to, however it gets made.
The artifact and its job
Before the sections, the standard: a teardown is not a research report, it is a decision instrument. Its reader is about to write battlecards, choose creative angles, set positioning or brief the board - and every section either feeds one of those decisions or wastes the reader's attention. This framing settles most arguments about what belongs: the rival's founding story belongs only if it explains current behavior; the complete feature matrix belongs only where features are the battleground. The competitive intelligence pillar's eight watching surfaces provide the raw material; the teardown is those surfaces synthesized against your situation.
The section map
| Section | Contents | Feeds |
|---|---|---|
| 1 · Identity | Firmographics, structure, stated positioning | Baseline; the entity everything else attaches to |
| 2 · Momentum | Funding, hiring velocity and shape, announcements | Threat sizing; where they are investing |
| 3 · Positioning | Claims, pricing posture, segments courted | Your positioning and battlecards |
| 4 · GTM motion | Sales model, partnerships, channels | Deal strategy; where you will collide |
| 5 · Paid presence | Live creatives, angles, formats, longevity | Creative angle decisions |
| 6 · Organic + AI footprint | Keywords held, content strategy, answer share | Content plan; GEO targets |
| 7 · Weaknesses | Exposed gaps, each with its evidence | The plays |
| 8 · Plays | Concrete moves, each naming its weakness | The next quarter |
Sections one and two: identity and momentum
Identity is table stakes done quickly: size, structure, ownership, the positioning they claim for themselves quoted from their own pages. Momentum is where the teardown starts earning: funding recency and size (a rival twelve months past a raise behaves differently from one twelve weeks past), hiring velocity and shape - the postings read from competitor tracking that shows which functions are growing, which is the closest public proxy for strategy - and the announcement stream. The section's summary question: is this company accelerating, coasting or retrenching, and where is the investment concentrating? Answered with evidence, that one paragraph reframes everything after it.
Sections three and four: the go-to-market read
Positioning reads their claims against the market's: which words they own, which segments their casing courts, how their pricing posture (published, gated, "contact us") signals their motion. The GTM section reconstructs how they sell - sales-led or product-led, partner channels, the events and communities they work - from postings (SDR hiring is a sales-motion tell), partnership announcements and observable channel presence. Together these two sections answer the question sales actually asks: when we collide in a deal, what will they say and how will they run it? That answer feeds directly into the sales battlecards system.
Sections five and six: the digital footprint
Paid presence comes from the Meta Ad Library read: active creatives, the angles they crowd, formats, and longevity as the performance proxy - the same data the creative loop consumes, here framed as "what is their paid strategy and where does it leave gaps". The organic and AI section maps keywords held, content investment patterns, and - the newest column - their share of the category's AI answers: which buyer questions the assistants answer with their name. A rival who owns the answers your buyers ask is winning a war some teardowns still do not have a section for.
The synthesis: weaknesses and plays
Everything above is evidence; these two sections are the deliverable. Weaknesses are exposed gaps with their evidence attached: the segment their pricing abandons, the angle their creative ignores, the buyer questions they lose in AI answers, the churn tell in their postings. No armchair psychology - only gaps an observable artifact demonstrates. Plays convert weaknesses into moves, each one naming its weakness, its owner-shaped action and its surface: "their feed has crowded the speed angle and left reliability open - test the reliability angle in next week's creative testing slot". A teardown whose plays could have been written without the evidence has a synthesis problem; a teardown without plays at all is a museum exhibit.
The citation rule
One closing note on freshness: a teardown this complete decays in weeks - which is why its mature form is the scheduled re-run with a highlighted diff, the pattern of the teardowns-on- a-schedule use case. The first teardown is the baseline; the machine is the cadence.
Frequently asked questions
What should a competitor teardown include?
Eight sections: identity, momentum (funding and hiring shape), positioning, GTM motion, paid presence from the ad libraries, organic and AI-answer footprint, evidenced weaknesses, and concrete plays that name the weakness each exploits.
What separates a good teardown from screenshots with adjectives?
Synthesis against your situation and the citation rule: every claim traces to an observable artifact, and the document ends in plays a team can run - decisions, not description.
How current does a teardown stay?
Weeks: creatives, postings, pricing and answer share all move on that clock. The mature practice is a scheduled re-run with a diff rather than an annual rebuild.
What is the most underused teardown section?
The AI-answer footprint: which of the category’s buyer questions the assistants answer with the rival’s name. It is measurable, moves weekly, and increasingly is the shortlist.
Can an agent produce a teardown this complete?
Yes - the surfaces are all observable and the synthesis is exactly what a mission with citation discipline produces: every finding carrying the tool output behind it, on demand or on a schedule.
Sources
- Meta Ad Library - the paid-presence evidence base
- SCIP - competitive intelligence ethics and practice
Every playbook on this blog ships as a runnable mission.
Open a workspace and the playbook library is waiting - describe the outcome and the agents carry it end to end, on your plan's monthly credits.