Infer the Equation. Solve for the Coefficients.

How a consultancy walks into the first meeting with a falsifiable theory of a prospect’s structural friction — compiled from industry priors and public evidence, never claimed as “biggest” until internal measures earn the word.

Scott Farrell · LeverageAI · Longer article (Stage A master for the full ebook)

The stronger first-meeting posture is not: “Tell me your biggest problems and I’ll suggest some AI ideas.”

It is: I have formed an evidence-backed theory about where your business model burns customer attention, senior labour and margin. Here is why I believe it, what would disprove it, and the product that becomes possible if it is true.

That sentence sounds like arrogance until you put two constraints on it. First, outside evidence may only nominate the problem. Second, only internal evidence may earn the word “biggest.” Between those constraints sits a compiler — an extension of the Proposal Compiler’s Marketplace-of-One motion — that turns industry structure and public company expression into a killable artefact: the Friction Thesis.

Why the organisation is a poor first witness

Structural friction is usually the last thing a firm can name cleanly about itself. It has become how the work “normally” feels. It is distributed across sales, architects, delivery and finance. Each function sees a fragment. Each fragment is described in the vocabulary of the current process.

Ask sales and you hear slow approvals. Ask architecture and you hear weak requirements. Ask delivery and you hear overselling. Ask finance and you hear margin variance. Each account can be correct while none names the system manufacturing all four outcomes.

The Friction Attack Surface already gave the outside observer a sharper question: where are customers or staff repeatedly translating stable intent into the interactions demanded by the provider’s systems and business model?¹ That friction may be accidental architecture. It may also be wrong friction created by the commercial model itself — the kind of friction that analytics sometimes celebrate as engagement.²

High-effort commercial paths are not a neutral style preference. Customer-effort research in the CEB lineage found that 96% of customers with a high-effort service interaction become more disloyal, versus about 9% after a low-effort experience.³ Consulting’s unpaid pre-sales and SOW dance is not a booking app — but the economic shape is cousin: stable intent forced through a high-effort translation machine that neither side fully trusts.

Industry first, company second, residual last

Vertical-of-One remains true: company-specific workflows, politics and exceptions decide AI fit. It does not require an empty page.

INDUSTRY STRUCTURE
  What friction normally follows from this market and business model?
        ↓
COMPANY EXPRESSION
  How does this firm’s public offer appear to instantiate or resist that structure?
        ↓
INTERNAL RESIDUAL
  What is genuinely different — scale, politics, unit economics, hidden systems?

That is the Seed Wiki move applied to commercial diagnosis: ship the skeleton; ingest the difference. Discovery becomes deviation hunting rather than blank-page interviewing.⁴

Build an industry friction seed

For data consultancies, the seed describes an expected transaction grammar:

Client experiences an outcome problem
→ translates it into a consulting request
→ salesperson qualifies
→ consultants reconstruct context
→ architects imagine a target
→ delivery estimates incomplete work
→ price is negotiated
→ SOW freezes a provisional model
→ delivery discovers the actual estate
→ changes and margin consequences follow

The seed also carries expected actors, artefacts, handoffs, pricing mechanisms, typical wrong friction, and candidate AI-constituted replacements. For each target, public evidence fills or challenges those stubs: website language, service catalogue, case studies, job ads, role mix, partnerships, executive posts, downloadable offers. A firm that describes everything as “tailored transformation,” advertises a large multidisciplinary bench, and has no narrow entry offer has already made a compressed statement of its commercial architecture — even without disclosing its internal proposal process.

Generate diagnoses before you search

Do not search “problems at Company X” and summarise the web. Generate competing structural diagnoses first, then run targeted research to attack each. That is reasoning-guided search: thought sets the research questions; retrieval does not set the agenda.⁵

For a mid-sized data consultancy, five candidates typically appear:

1. Unpaid proposal and SOW production is the major friction.
2. Poor reuse of client and project history.
3. Absence of narrow, buyable entry products.
4. Downstream margin variance from weak current-state evidence.
5. Installed base without systematic next-offer matching.

The winner survives on structural inevitability, independent public signals, plausible economic consequence, a named buyer, an accessible first test, and a credible replacement product hypothesis. The losers stay in the document with the reason each lost. That is the John West proof of judgment — the rejections that make the surviving call credible.⁶

The lead thesis is structural, not symptomatic

“They spend too long writing proposals” is a symptom target. The structural thesis for a mid-sized data consultancy usually looks more like this:

The firm’s exposed commercial friction is that the first major transaction asks the client and consultancy to agree on the price of a bespoke transformation before either possesses a shared, evidence-backed model of what must be transformed.

In that structure the senior SOW author becomes the manual join across client request, account history, prior delivery, architecture, staffing and commercial risk — the join algorithm of the commercial process.⁷ A fixed-price readiness / preparedness product does not merely accelerate proposal writing; it changes the first commercial unit from uncertain transformation to purchased shared evidence. What that product becomes is owned elsewhere — the successor-offer object and a worked preparedness family.^8,9

The 15-field Friction Thesis

The compiler’s output is not a paragraph of clever insight. It is a structured artefact:

The first meeting changes purpose

You do not open with “What keeps you awake at night?” You do not announce “I diagnosed you from your website.”

You say: I have formed a hypothesis from the structure of your market, your public offer and the economics of bespoke data consulting. I may be wrong. Here is the evidence, the alternatives, and the small set of internal measures that would confirm or overturn it.

The client’s job is no longer to create the insight. Their job is to supply the local residual that confirms, modifies, exceptions, or kills the thesis. Strategy Engine doctrine already states the commercial rule: claims about you carry receipts; claims about them carry a test.¹⁰

The minimum validation pack for the data-consultancy thesis is bounded — not months of interviewing. Roughly nine internal observations: unpaid hours by role into an average proposal; win rate of qualified proposals; discounting after initial scope; estimate-versus-delivery match; recurring change requests; gross-margin variance by engagement type; time from serious discussion to paid start; share of opportunities that could fit a bounded readiness product; and whether clients would pay a modest fixed amount to reduce uncertainty first. If those observations do not show material pre-sales cost, scope variance or client hesitation, the thesis weakens. That is success: you killed an attractive but incorrect proposal before you built around it.

Commercial systems that invest in evidence before the formal proposal already show better economics. Boutique consulting commentary treats proposal-to-win rates around 20–30% as a red light, with much higher rates when the sales process is narrow and value has been traded before the document ships.¹¹ Segmented RFP analysis in one published vendor case pattern found roughly 48% wins on qualified enterprise deals with pre-RFP discovery versus 28% without — and far worse on low-fit inbound.¹² Those numbers do not prove your thesis about any one firm. They prove that arriving empty is an expensive habit.

The asset is the library

The scalable asset is not a list of target companies. It is a library of industry friction seeds and observed company deviations. After several data consultancies, the prior knows the common SOW transaction, where senior labour enters, which public signals predict productisation weakness, which metrics validate the diagnosis, and what conditions killed earlier theses. Each new target starts from a stronger skeleton. Marketplace of One already said: research before engagement, apply your kernel, synthesise a distinctive recommendation, make the artefact the first proof.¹³ The upgrade is that the artefact now proves something sharper than “I understand your company.” It proves: I can detect a structural problem you had normalised, distinguish it from nearby symptoms, propose a different commercial mechanism, and make my diagnosis easy to falsify.

Do not ask prospects to diagnose themselves. Bring them a falsifiable theory of where their business burns attention and margin. Let their internal evidence supply the delta.

You do not need the client to see the car before you do. You need enough humility and machinery to prove whether the car belongs on their road.