Market attractiveness analysis

Score market attractiveness across five dimensions, each 1 to 5, where 5 is always the favorable end and 1 is always the unfavorable end. Watch the direction if you've come straight from the Five Forces. The intensity scale you learned there runs the other way, with 5 meaning a severe, profit-destroying force. Here, on both axes, higher is always better. That's a deliberate choice, so you can read market and position on the same plane without a sign flip.

How do you score market attractiveness?

Add up the five scores (range 5 to 25) and read the total as a band, never a point estimate: ≤11 unattractive, 12 to 18 mixed, ≥19 attractive. The banding matters more than the arithmetic. The Porter lesson made the same point about the five-forces index ("the sum is not the output"), and it applies twice as hard here, because a five-dimension composite invites the kind of false precision that turns a judgment tool into a spreadsheet trap.

So what you hand over is three sentences, never a figure like "17.4": the band, the dominant driver (the one or two dimensions actually doing the work), and what would have to change to move the band. Why three sentences and not a number? Two markets can land in the same "mixed" band for completely different reasons. One scores evenly across all five dimensions. Another has a spectacular growth rate dragged down by poor structure and a thin, shrinking pool share. Only the three-sentence version tells a client which is which. The worked examples on the fragmented link and the dominant player land in nearly identical bands for that very reason, and the near-match is deliberate.

Name two traps before you score anything. First, growth is one-fifth of the score. A headline CAGR is the most seductive and least sufficient piece of evidence in a market-attractiveness case, because it's the easiest number to find and the easiest one to lead with. Second, score the profit-pool dimension at the correctly defined link. Use the five-step method you already own from the profit-pools work: define the chain, size total revenue, estimate each link's revenue and margin from proxies, then compute each link's share of the pool. Nothing here changes that method. The one addition is reading its direction forward, which the next part takes up.

Reading profit-pool migration before it shows up in the trailing numbers

The profit-pools work already named the final, most-often-skipped step in mapping a chain. Once you've computed where the pool sits today, ask which way it's migrating, because today's shape is a snapshot of forces that are still moving. Here that one-line prompt becomes a repeatable diligence habit. Why bother? By the time a shrinking pool share shows up in a target's own trailing P&L, the price you'd have to pay to exit that position has usually moved against you already. The skill is underwriting migration before it's visible in the numbers you can pull from a filing. In practice that means reading indicators that move earlier than revenue and margin do.

Watch five leading indicators. They usually turn in this order:

1. Capital allocation across the chain. Capex and announced-capacity data across every link, and not just the target's, tell you where the industry itself is betting its money. Often the earliest visible signal is new capacity piling into one link while another link's incumbents merely maintain existing plant. Capital moves before revenue does.
2. Patent and R&D-filing mix. Watch for a rising share of chain-wide patent filings or R&D headcount in one link versus another. It's a multi-year-ahead signal of where the technical frontier is moving, and eventually the pricing power with it.
3. Regulatory and policy direction. Mandates, subsidies, safety standards, and tariff walls reliably favor one link over another. Treat the direction of policy as a signal. Before you quote a number from any specific incentive or mandate, check it against the regulator's own text. A headline isn't a source.
4. Customer-side sourcing behavior. Notice when customers change what they ask for in RFQs, qualification criteria or contract terms, and start specifying the upstream or downstream attribute more than the link you're scoring. That's a demand-side vote, and it usually comes several quarters before a visible margin shift.
5. Talent flows. Engineers, scientists, and specialist hires moving from one link's employers toward another's give a slow signal, but a very hard one to fake. People bet their careers on where the interesting (and eventually well-paid) work is going.

Distinguish this from cyclical share churn, which looks similar in a single trailing chart but is not structural: a link's margin falling in a downturn alongside every other link's margin falling in the same downturn is the industry-wide cycle, not a pool migrating away from that link specifically; a link's margin falling while other links in the same chain hold or improve, in the same period is the structural signal this section is built to catch. The test is always relative, score the link against its chain neighbors in the same period, never against its own history alone.

{
  "skill": "Surfacing candidate leading indicators of profit-pool migration before they show up in trailing financials",
  "use": "Have a chat or research assistant draft a first-pass list of migration signals to check for a specific value-chain link -- capex announcements, patent-filing mix, regulatory direction, customer RFQ language, talent flows -- so you know exactly where to point primary-source research, never as the finding itself.",
  "tools": ["Chat assistants -- Claude, ChatGPT", "Research/agentic tools -- a browsing-enabled assistant for scanning capex and capacity announcements across a chain", "Document/PDF AI run over a set of MD&A sections or expert-call transcripts you already hold"],
  "prompt": "For the <CHAIN> value chain and the <LINK> link specifically, list the five most likely leading indicators that its profit pool is migrating toward or away from <LINK> over the next 3-5 years, and for each one name the specific primary source I should check myself (a filing's capex table, a patent-database search, the regulator's own published text, a dealer or distributor data point) to verify the signal.",
  "verify": "Every migration signal an assistant proposes is a hypothesis, not a finding. Trace each one to the named primary source before it enters your scorecard, and Recompute any pool-share arithmetic yourself from the underlying revenue and margin data -- a plausible-sounding migration story is exactly the confident, unverified claim the Primary-Source Guardrail exists to catch (Guardrail 1-5, especially #1 Trace and #3 Recompute).",
  "diy": "You must already be able to build the five-dimension market score and run M4.03's five-step profit-pool method by hand. AI only accelerates which signals to go look for; it does not replace scoring the dimensions or running the method yourself.",
  "market": "IN"
}

Scoring competitive position: five dimensions, one honest band

You score position for a specific player, never for "the market". The mechanics match attractiveness: five dimensions, 1 to 5 each, 5 always favorable, summed to a 5 to 25 band with the same cut points of ≤11 weak, 12 to 18 average, ≥19 strong.

What mistakes do people make when scoring position?

The first dimension carries the rule people break most often in this scorecard: score the trend, never the level alone. A snapshot share of 22% and second place in the market can describe a durable, stable position or one in active collapse. On the day you pull the number, the two look identical. The share-stability test exists because share churn, or its absence, over years is one of the cleanest empirical signals of position durability there is. Reuse it here as it stands, and never accept a single-period share figure as a finished answer. The worked example on why a snapshot share number lies is built entirely around this trap.

The fourth and fifth dimensions lean hardest on tools you already own. Customer economics comes straight from the business-model teardown's contribution-margin verdict, read as a position signal. Differentiation comes straight from the moat numbers test and the moat grammar, used whole. If you catch yourself re-deriving the five moat sources or re-running the Greenwald collapse from scratch inside a position score, you've drifted back into work those earlier lessons already did.

The plane: four quadrants, four verdicts: and why this is not the BCG matrix

Plot the market-attractiveness band on one axis and the position band on the other, and four corners fall out, each with its own correct advisory action:

Real engagements rarely hand you a clean corner, and two things happen far more often. First, most real targets land in the "mixed" middle on one or both axes. When they do, the other axis usually does most of the work in deciding the call. The worked examples on the fragmented link in a hot category and the dominant player in an unglamorous market both score their markets in the mixed band, at 14/25 and 15/25. Those are nearly identical, yet the two land at opposite verdicts, purely because their position scores are almost polar opposites (7 versus 22). Market scores usually cluster in the middle, and when they do, position decides the outcome. Hold on to that more tightly than to any single quadrant label.

How is the plane different from the BCG matrix?

Second, the plane looks like the BCG growth-share matrix from the portfolio-frameworks lesson, so name the family resemblance and then the real difference. The BCG matrix plots two observable numbers on a fixed grid: a business unit's market growth rate, and its market share relative to the largest rival. It's built for allocating capital across one company's own portfolio of businesses at a glance. Its blind spots (experience-curve dependency, market-definition sensitivity, unit interdependencies, cost-of-capital blindness) are documented in full in that lesson, and you won't repeat that work here.

The plane you're building plots two independently built, five-dimension, evidence-scored composites. It's made for diagnosing one specific target relative to its market. That target might be a division you already own, a company you're diligencing for acquisition, or a market a client is asking whether to enter. So the BCG matrix is a fast, standardized snapshot for portfolio triage. Your scorecard is the slower instrument, graded on filings and expert calls, that you build when the capital at risk on one decision is large enough to earn it. That's the commercial-due-diligence setting the CDD toolkit sends you into next.

Why score the forward quadrant as well?

One last point separates a diligence-grade read from a casing-grade one. Quadrant membership is not a fixed address. A target in the compounder corner today can be drifting toward the trap corner. It happens when its position's fifth dimension, the capacity to capture a moving pool, is deteriorating while its share and margins still look fine this year. The worked example on gating an acquisition and pricing a forward risk builds that case. So score today's quadrant and the migration-adjusted, forward quadrant side by side, and price or structure a deal against the forward one. An acquirer who pays a compounder-multiple premium for a target already sliding toward the trap corner has bought yesterday's business at tomorrow's-problem's price. Most often it happens because nobody did the profit-pool migration read described earlier.