%202.png)
Competitive and Market Research
Build a defensible view of a market and its players from external sources, with every source ranked for reliability and every unverifiable claim labelled.
Free. Works on Claude, ChatGPT, Gemini or any assistant that accepts a skill file.
What this skill does
The method, encoded.
In competitive and market research, most available evidence was published by parties with an interest in the answer. Four failures recur. Marketing claims are read as findings, because they arrive in the register of fact. Market sizes get quoted without anyone establishing what they count, and most widely cited figures collapse to a single unverifiable estimate that other sources repeated. Competitors' stated intentions are treated as evidence of what they will do. And the market is described entirely from the supply side, without anyone establishing what buyers actually buy.
This skill supplies the working method. It ranks every source on an explicit six-tier reliability hierarchy, from mandatory filings down to unverifiable material that carries no claims at all. It separates what a competitor says from what it does, treating recruitment, pricing changes and filings as the harder evidence they are. It requires any published market size reconstructed on seven attributes before use, and any figure traced to its origin so citation chains collapse. It assesses the market from the customer side, where the largest competitor is frequently doing nothing at all.
And it is explicit about the boundary: what competitive research cannot establish, and the methods it must never use.
Best used for
- Sizing a market or segment from secondary sources with no primary budget
- Building a competitive landscape for an entry, launch or investment decision
- Testing whether a widely cited market figure is defensible
- Establishing what competitors have done rather than announced
- Identifying the real substitutes for a proposition
- Providing external context for a business case that will be scrutinised
Typical inputs
What you give it.
The decision the research serves and the precision it requires, A market or category definition stated as an operational rule, Access to at least one primary record source (registers, filings, regulator data), Optional internal win-loss, pipeline or transaction data, Optional figures already circulating internally, with their believed origin, Optional assumed competitor list to be tested, Optional sector regulatory or statistical returns
Typical outputs
What you get back.
Source table with reliability tier, date of data and traced origin per source, Competitive set defined by substitution, including non-consumption, Says-and-does evidence base per competitor, with divergences noted, Citation chains collapsed to their true number of independent origins, Reconstruction note per market size figure across seven attributes, Top-down and bottom-up size estimates with a stated range and sensitivities, Customer-side view of buyers, share of requirement and switching, Competitor claims labelled with incentive context and definitional gaps, Statement of what cannot be established about economics, performance and intentions, Claims register with a status per claim, and a stated shelf life
Method coverage
What the skill works through.
- Why competitive evidence needs a reliability hierarchy
- The six tiers, from mandatory filings to unverifiable claims
- What a competitor says versus what a competitor does
- Recruitment and pricing as early signals
- Tracing a market figure to its origin, and collapsing citation chains
- Reconstructing a published market size before you use it
- Building a size estimate top-down and bottom-up
- Assessing a market from the customer side
- Reading competitor performance claims against their incentives
- What competitive research cannot establish
- Ethical and legal boundaries in competitive intelligence
Download
Free skill. One file.
Enter your email once. Every skill you download after that takes a single click.
How to install
Add the skill file and the five kernel protocols to a Claude Project, a ChatGPT Project, a Gemini Gem, or paste them at the top of any assistant conversation. Then give it your real research material, not a description of it.
Download skillQuestions
Common questions.
How do I verify a market size figure?
Reconstruct it before using it. Establish seven things: the universe being counted, the unit and its recognition basis, the inclusions and exclusions, the geography, the base year and whether it is actual or forecast, the method, and the assumptions the model rests on. Then trace the citation backwards until you reach a primary record, a stated method, or a dead end. A figure you cannot reconstruct does not enter the analysis as a number.
Why do published market sizes disagree so much?
Usually because they count different things, not because one is inaccurate. Different universes, units, inclusions, base years and recognition bases produce figures that differ by multiples. Reconstruct both before deciding either is wrong; the reconciliation is often possible and is more informative than picking one.
What is the citation chain problem?
Four sources appear to corroborate a figure, but each cites another and all terminate in a single original estimate, so repetition reads as consensus. Collapse the chain by tracing every figure to its origin, and report "traces to a single estimate published by X in year Y, method not disclosed" rather than "widely reported".
How reliable are different competitive intelligence sources?
They vary enormously, and reliability does not correlate with findability. In descending order: mandatory filings, registers and regulator data; your own dated observation of published prices, terms and job advertisements; unaudited company communication such as investor decks and press releases; independent third-party commentary whose own source can be traced; marketing material, which is evidence of positioning only; and unverifiable material, which carries no claims at all.
How do I tell what a competitor is really doing?
Separate says from does and keep them in different columns. Says covers stated strategy, roadmap and positioning. Does covers prices actually charged and when they changed, products actually available, markets actually served, filings made, capital raised, and roles being recruited. Recruitment is the most underused signal: capability is hired six to twelve months before it appears in a product.
Can I work out a competitor's revenue or margins?
Not from outside, and estimating them from headcount, footprint or category benchmarks produces a number that will be quoted without its derivation. Unit economics, margin, acquisition cost, churn and true customer counts are outside the reach of external research. Independent public traces can bound a competitor's scale as a stated range, but only if the traces are shown rather than blended into an estimate.
What are the ethical limits of competitive intelligence?
Misrepresenting your identity, employer or purpose to obtain information is prohibited, including posing as a prospective customer, job applicant, student or partner. So is inducing a breach of confidence, approaching a competitor's staff under a pretext, and using material you know to have been improperly obtained. Direct exchange of commercially sensitive information with competitors carries competition-law exposure and is a matter for legal advice, not research judgement.
Who are my real competitors?
Define the set by what buyers substitute rather than by who describes themselves the way you do. That usually adds three things a supplier-drawn landscape misses: doing nothing, doing it internally, and an adjacent product being used for the purpose. In many categories the largest competitor is non-consumption, and it never appears on a vendor list.
The skill chain
Works well with.
Research where people already are.
Analyse it where you already work.
Yazi helps researchers conduct surveys, AI interviews and longitudinal research directly through WhatsApp.
