Full SKILL.md

Market Competitive Intelligence

Builds an industry fact base with graded evidence — profit-pool map (Gadiesh and Gilbert), triangulated market sizing, Porter's Five Forces rated from indicators, competitor profiles, and client implications. Use whenever a strategy question turns on where profit sits in a value chain, how big a market is, or what competitors will do, or the user says "where is the money in this industry", "map the profit pool", "is this an attractive market", "run five forces", "size this market", "why are our margins so thin", "competitor fact base". Not for synthesising supplied internal documents (document-synthesis), interview programmes (primary-research-synthesis), scenarios and rival response (war-gaming), or choosing among options (strategic-choice).

Produces the fact base that strategic-choice, war-gaming and product-strategy Phase 1 consume: where profit sits, why, who is positioned to take it, and what that opens or closes for the client. Sources and the full Porter driver lists are in references/frameworks.md; read it only when a rating needs defending.

When not to use

  • Open-ended multi-perspective external research: boardroom-storm-research, then bring its output here as graded inputs.
  • Supplied internal documents: document-synthesis. Interviews: primary-research-synthesis.
  • Scenarios or competitor response modelling: war-gaming (this skill supplies its fact base).
  • A product-specific market attractiveness scoring inside a product strategy: product-strategy owns the scoring; call this skill for the pool map and sizing.

Inputs

Ask for, or infer and grade: the decision and decision-maker; the client's own numbers (revenue by segment, margins, top-customer and top-supplier concentration, win-loss, renewal rates: the cheapest A-grade evidence, and the most often forgotten); a first view of the boundary (product, customer, geography); public sources to hand; expert or customer access; time and required accuracy (±20% for a board discussion is fine; an acquisition price is not).

Proceed on graded assumptions rather than stall. If external gathering is a large job, delegate a precise list of needed numbers to boardroom-storm-research and continue.

Workflow

  1. Evidence standard. Open a source ledger (references/templates.md §1), one row per number: value, source, date, grade, analysis fed. Grades: A hard primary or audited data; B triangulated from two independent sources or several consistent experts; C single source or reasoned inference; D opinion, including management's view. Name the three numbers that, if wrong by 30%, change the recommendation; they must reach B before presentation.

  2. Question and boundary. One-sentence decision question. Arena on three axes: products by function not catalogue; customers by buyer groups with distinct economics; geography where the same competitors meet the same customers. Pool is drawn one level wider than the forces (whole value chain vs the client's activity). Tests: do participants inside face the same forces? Is the client's actual bid competitor set inside? Done when the boundary statement has in/out lists and names the two or three boundary choices that would change the answer.

  3. Value chain. Six to twelve activities from raw input to end use, including side activities (financing, insurance, service, logistics, data) and activities the client does not perform. For each: performers, revenue basis, data availability. Test: could a competitor profit from an activity not listed?

  4. Profit pool, four steps. (a) Define: operating profit before interest and tax, normalised year, confirmed boundary. (b) Total pool two ways: end-customer revenue × blended margin, and sum of activities; record both and the gap. (c) Per activity: revenue and margin, allocated in order of preference from pure-play proxies, segment disclosures with overhead on revenue, sampled private players (sample size, median, weighted mean, range), split multi-activity firms by their own mix. Draw bars: width revenue, height margin, area profit; tabulate share of revenue vs share of profit. (d) Reconcile activity sum to total, check the value identity along the chain, run sensitivity on the one or two margins that move the total, mark C-grade activities, add five-year trend per pool with driver. Done when each pool has one sentence on why it is deep or shallow (a hypothesis for Stage 5). Test: if the map merely confirms what management believes, check it was not pre-shaped.

  5. Size the market that matters, three ways. Define what is sized in one sentence (revenue of whom, for what, to whom, where, which year). Top-down from a verified aggregate with share factors; bottom-up from unit counts × adoption × spend × share × yield; comparable from an analogous market ratio, naming the structural point where the analogy may fail. Every factor in the ledger with grade. Reconcile gaps by naming the factor, never by averaging. Report point estimate, range across methods, growth rate with sources; size profit as well as revenue where margins differ. Done when methods agree within 15–20% or the disagreeing factor is named.

  6. Five forces from indicators. Run on the client's activity as bounded. For each force, evidence Porter's drivers with at least one measurable indicator (value, source, grade) before rating 1–5. Anchor the scale: 1 ignorable in price or cost targets; 3 constrains part of the business; 5 sets the margin ceiling for the whole activity. A 4 or 5 must trace to a number in the client's own P&L (discount given, pass-through missed, bid lost on price). Add trend and cyclical-vs-structural. Growth, technology, government and complements are factors acting through forces, not forces. Effects (low margins, price sensitivity) are what forces explain, not evidence of them. Done when the forces table explains the pool map; a deep pool under hostile forces means one analysis is wrong.

  7. Competitor fact base. Three to six: scale leader, closest bid rival, fastest-growing entrant (including different models), any firm already in a pool the client wants, the supplier or customer most likely to integrate. Profile each (references/templates.md §7): ownership, size, growth, margin, footprint, customers, proposition, cost position, capabilities, recent moves, stated goals, apparent assumptions, likely next moves, every number graded. Close with a one-line likely reaction to each client move under consideration. Done when each competitor's driving belief about the industry can be stated in one sentence.

  8. Implications. In order: where profit sits and how it is moving (two sentences with numbers); why (binding forces with strength and grade); where the client sits (share of revenue vs profit, margin vs pool average, exposure); who is positioned to take the pools the client wants; two to four options the evidence opens or closes, each mapped to one of Porter's three responses (position where forces are weakest, exploit a change under way, reshape a force) with the condition under which it works. Do not choose; state which hypotheses the evidence supports or refutes, and the gaps that would most change the picture.

Output

Headline conclusion in the first 150 words. Then, each once, 1,500 to 3,500 words including tables:

  1. Fact-base summary (one page, references/templates.md §9)
  2. Boundary statement
  3. Profit-pool table and chart specification (activity, revenue, margin, profit, share of revenue, share of profit, trend, grade) with reconciliation note
  4. Market sizing sheet (three methods, reconciliation, point, range, growth)
  5. Force-by-force evidence table (drivers, indicators, values, sources, grades, strength, trend)
  6. Competitor profiles and comparison table
  7. Options with conditions; hypotheses supported or refuted; evidence gaps
  8. Source ledger as an appendix

Assumptions follow the conclusion they affect. No method vocabulary, checkpoint labels, framework authors, skill names or the worked example's client in the deliverable. Full example: references/worked-example.md.

Quality bar

Run silently; fix, do not report.

  1. The three decision-critical numbers are grade B or better, or the summary says they are not.
  2. Activity pools sum to the total pool within the stated sensitivity.
  3. Share of revenue vs share of profit is tabulated for every activity.
  4. Sizing methods agree within 20% or the disagreeing factor is named.
  5. Every force rating of 4 or 5 traces to a client P&L number.
  6. Forces explain the pool map.
  7. Each competitor has a stated goal and a stated assumption, not just a description.
  8. Each option maps to one of the three structural responses and has a condition.
  9. No number without a ledger row.
  10. Headline in the first 150 words; no leaked method vocabulary.

Anti-patterns

  • Boundary drawn to fit the answer.
  • Revenue map presented as a profit map.
  • Assigning a multi-activity firm wholly to the activity it is known for.
  • Market sized by averaging three methods.
  • Force rated by adjective before the drivers are evidenced; effects used as drivers.
  • Competitors described by last year's moves rather than goals and assumptions.
  • A fact base that ends without implications.

Hand-offs

References

  1. Frameworks: the intellectual foundation in depth

    3,948 w
    references/frameworks.md6 sub-sections
  2. Templates and checklists

    1,854 w
    references/templates.md11 sub-sections
  3. Worked example: Northgate Industrial Supply

    4,143 w
    references/worked-example.md9 sub-sections