Creating the Research Spine

The Cost of a Soft Foundation

You have almost certainly sat through a presentation built on a market number nobody in the room could source. It was probably cited correctly, which is exactly what makes this hard to catch.

A weak industry foundation doesn’t announce itself. The work it produces is careful, well sourced, internally consistent, and describes the wrong world.

Everything downstream makes interpretive decisions. Opportunity Discovery identifies the Strategic Truths. Brand and Positioning turns those into language. Audience Architecture determines who matters and in what order. Journey Mapping establishes what each audience has to see, understand and believe before it will move. Every one of those decisions is only as sound as the industry reality underneath it, and not one of them will announce the problem, because each will be perfectly consistent with the deliverable above it.

The Research Spine gives every downstream recommendation a factual boundary it has to respect. That’s a smaller job than strategy, and a harder one to skip.

How the Work Gets Done

Establishing the Industry Boundary

Start with the most authoritative classification available, usually anchored in federal definitions like the North American Industry Classification System. Then we identify the adjacent industries most often confused with the category and say why each one is in or out.

That definition controls the rest of the research. Market size, firm counts, employment, revenue estimates and growth rates aren’t allowed to drift between incompatible categories because another number is newer or more impressive. If the definition changes, the numbers change with it.

This is the least glamorous work in the engagement and the most consequential. Build a strategy on a market that exists only because unrelated categories got added together, and the precision of everything that follows is largely cosmetic.

Test your own numbers before anyone else does. Take the market figure your firm quotes most often and trace it back to the classification behind it. If nobody can name the code, the survey or the year, the number is decoration and it has probably been in a proposal.

Sizing and Shaping the Industry

With the boundary set, determine the defensible size and economic shape of the industry. Prioritize current federal data, authoritative industry associations, and sources that can explain what their figures measure. When credible estimates disagree, go at their definitions, dates, methodologies and underlying populations rather than averaging incompatible numbers into something that looks precise.

The useful output is a picture of the category, not a number for a slide. Firm and establishment counts, employment, revenue, geographic concentration and ownership patterns, and then the economics underneath the totals. Where margin comes from. What constrains capacity. Which parts of the industry behave nothing like the aggregate.

When two credible sources disagree, the productive move is to compare what each one counted rather than split the difference. Averaging incompatible definitions produces a number with no referent, and it will look more precise than either source it came from.

What the rest of the engagement inherits is a commercial environment it can defend, down to which figure came from where.

Understanding How Work Gets Awarded

An industry is more than its financial totals. Across nineteen sections the research also moves from definition and history through economics, competitive composition, buying behavior, workforce, technology, regulation and outlook.

Document who buys the service, who influences the selection, how firms compete, what constrains capacity, and which forces decide how work gets awarded. This is still industry research. None of it decides which buyers a client should pursue, or declares a competitor vulnerable. It establishes the decision environment so the client-specific work doesn’t begin with assumptions disguised as insight.

Decision Architecture depends on that distinction. Relevance, Credibility, Proof, Safety and Access behave differently depending on where the buyer is standing. In a category where a bad selection ends someone’s career, Safety costs more to clear and arrives later than anyone plans for. In a fragmented market with no dominant name, Credibility does more work than Proof. Those conditions belong to the industry. Map them before designing anything meant to move a buyer through them.

Separating Change From Noise

Technology, regulation, labor conditions, capital markets and broader economic forces can change an industry well before the change shows up in anybody’s marketing. Examine those forces while holding current evidence apart from prediction. An emerging technology and a widely adopted one are different facts. So are a proposed regulation and an enacted one, and marketing material collapses that distinction constantly.

The outlook keeps three categories apart: what has happened, what appears to be developing, and what’s still uncertain. The work downstream has to account for change without pretending the future already arrived.

Building the Evidence Library

The foundation closes with a statistics library and a complete source index. The statistics library puts consequential figures where later deliverables can find them without reconstructing the underlying research. The source index records where each fact came from, what it measured and how current it is.

Any firm can start this today with one spreadsheet. Every figure the firm uses in public, where it came from, what it measured, and the date it was published. Most firms find at least one number nobody can source, and it is usually the one on the homepage.

That’s what protects the system from drift. Someone remembers a statistic without remembering its definition. It goes into a presentation, then into a proposal...and eventually the claim becomes familiar enough that nobody remembers it was never quite accurate. Later deliverables can interpret the evidence. They can’t quietly change what it says.

What the Research Spine Does Not Do

It doesn’t identify the client’s position, recommend an audience, write a message, or declare that some industry condition represents an opportunity. Those calls need firm-specific evidence: completed work, client testimony, credentials and approvals where the category requires them, and an honest read of what the firm can’t do.

Keeping those decisions out of Process 00 protects both sides. Industry facts don’t get bent toward a preferred strategy, and strategic recommendations don’t get presented as the inevitable conclusion of neutral research. Anyone who has spent time in this business has watched a market number get quoted back inside a positioning argument as though it settled something. It never does, and by then nobody remembers who ran the number.

What Should Be Produced

A complete Research Spine produces:

• An authoritative industry definition and documented category boundaries
• A defensible account of market size, structure and economics
• Analysis of competitive composition, buyers and purchasing behavior
• Examination of workforce, technology, regulation and macroeconomic drivers
• A qualified forward outlook that separates evidence from forecast
• A statistics library and complete source index

The result is the canonical Vertical Intelligence Foundation for the category. Every deliverable after this one inherits it rather than researching the industry again from scratch.

Where AI Earns Its Place

AI has made a real difference in how fast sources can be located, classified and compared. It surfaces conflicting estimates, organizes large bodies of public information, and it’s good at revealing when six reports are quietly repeating one original source.

It can’t make incompatible definitions compatible, and it can’t turn a widely circulated estimate into a reliable fact. It will repeat a circulated error faster than any human researcher if nobody set the rules for what counts first. So we use it to widen the inquiry and carry the mechanical load of organizing evidence. The controlling facts still trace to identifiable sources, and the judgments about boundaries, comparability and uncertainty stay visible and stay ours.

What Process 00 Hands to Process 01

Opportunity Discovery is where the specific firm enters. Process 01 examines the firm’s market condition, its buyers and competitors, its referral relationships, its capabilities and the evidence it can produce, then sets all of that against the industry reality established here. Doing them in this order isn’t a scheduling preference. A firm’s opportunity is a claim about its industry, and a claim about an industry is worth exactly as much as the boundary it was measured inside.

Process 00 establishes what’s true about the industry. Process 01 determines where, inside that reality, the firm has a credible opportunity to matter.

Related reading

Decision Science: Expertise Pattern Recognition and Judgment. Explore the reasoning behind this part of the process.

How-To: How to show expertise instead of claiming it. Apply a specific part of the work.

Continue with Process 01.