Supplier intelligence: converting qualified opportunity into expected contribution

01 — Thesis: opportunity intelligence must alter a commercial decision

Intelligence is valuable to an OEM, EPC, equipment distributor or industrial-services supplier only when it changes an action: select a better target, decline an uneconomic bid, price risk correctly, enter an attractive segment or choose a defensible partner. An installed asset is not a procurement opportunity, a news announcement is not a funded tender, and a target company's name is not an introduction or an order. LARGITAS should distinguish the scale of an industry from accessible, eligible, funded and winnable packages.

A credible return-on-intelligence analysis measures avoided commercial pursuit expense or improved probability-weighted contribution, then subtracts the true cost of qualification, proposal preparation, delivery risk and the advisory fee. Demonstrations of this method are scenarios unless they can be supported by client CRM history and invoicing/collection records.

02 — Measured market context and its limits

Proyectos México's electricity profile reports approximately 111,195 km of electricity transmission, 908,195 km of distribution and 49.8 million electricity users at the end of 2025. Those are installed system metrics, not a pipeline of serviceable annual contract volume. They do not disclose replacement cycles, open invitations, allocated budgets or award probabilities. National energy demand and capacity growth alone cannot identify which supplier has a profitable route to a particular purchase decision. [S1]

Sectors adjacent to these networks include transformers and protection systems, control technologies, fiber networks, cooling systems and EPC/commissioning contractors. Each has different purchaser characteristics, technical qualifications, import content, insurance, warranty and cash-collection profile. Do not blend them into one conversion-rate funnel.

03 — Five-gate qualification hierarchy

• Gate 1 — project existence: identify the real project sponsor, asset owner, geography and technical requirement; reject duplicated press stories and false project stages.
• Gate 2 — economic need: determine what bottleneck the proposed equipment solves, quantified capacity/customer demand and credible funding rationale.
• Gate 3 — lawful procurement access: confirm public/private procedure, potential vendor eligibility, non-conflicted routes and decision timeline.
• Gate 4 — credible supplier fit: check certifications, lead times, technical specs, track record, capacity, warranties and downstream service obligations.
• Gate 5 — contribution and cash: model bid cost, win probability, gross-to-contribution margin, working capital, performance bonds, delay liquidated damages, taxes and collection risk.

A favorable score in the first four gates is not sufficient if the fifth gate is negative. A supplier should not pay to chase volume whose risk-adjusted contribution destroys value.

04 — Expected contribution model

Take a strictly hypothetical US$2m contract with 28% contribution after incremental delivery costs, giving US$560k contribution if won. Without enhanced intelligence, suppose an 8% chance of winning and US$15k commercial pursuit spend. After a documented qualification intervention, assume a credible 18% win probability and US$25k total pursuit spend. The two probabilities are assumptions—not causally established outcomes, observed win rates or LARGITAS performance claims.

• EV(base) = 0.08 × US$560,000 − US$15,000 = US$29,800.
• EV(qualified) = 0.18 × US$560,000 − US$25,000 = US$75,800.
• Incremental EV = US$46,000 before advisory cost and any incremental delivery risk.
• Extra pursuit spend = US$10,000.
• Break-even probability uplift = US$10,000 / US$560,000 = 1.786 percentage points.
• After a hypothetical US$12,000 research fee, modeled decision EV improvement is US$34,000, provided the estimated probability improvement is genuinely causal and no other costs arise.

The last line is a conditional expected contribution effect, not US$34k of revenue, realized profit, forecast precision or a proved return from intelligence. The fee is a scenario input, not a published LARGITAS rate.

05 — Why win probabilities are difficult to prove

A CRM uplift after qualification may reflect selection rather than better information. Stronger deals may receive more research and also be easier to win for other reasons. Correct for survivorship bias, opportunity aging, territory, incumbent status, contract size, sales cycle, technical fit, risk class and proposal quality. When sufficient data exist, compare matched cohorts, holdout groups or pre-specified intervention windows, and report interval uncertainty rather than a point estimate alone.

The model should be calibrated by sector, buyer class and stage: discovery → qualification → bid invitation → proposal → shortlisted → award → executed contract → cash collection. Treat stage transitions as conditional probabilities and exclude lost bids and withdrawals from an artificially favorable denominator. One US$2m opportunity cannot validate a population-wide 10 percentage-point win-rate improvement.

06 — Decide whether to bid, not merely how to bid

A no-bid recommendation can be the highest-value output. A US$40k pursuit with 3% probability and US$600k contribution if won has expected value US$18k − US$40k = −US$22k. Avoiding that bid preserves US$40k of planned spend, but the incremental economic improvement relative to an informed no-bid baseline must not double-count the foregone hypothetical award. A better opportunity might produce greater expected value even if its headline contract size is lower.

Scale the analysis into a constrained sales-capacity portfolio: choose binary bid decisions x_i ∈ {0,1} to maximize Σ x_i [p_i × contribution_i − bid_cost_i − expected_loss_i] subject to salesperson hours, engineering capacity, bid bonds, equipment lead time and working-capital limits. Then stress correlation: several bids may depend on the same government budget or single OEM shortage.

07 — Price, margin and working-capital bridges

Contract amount is not profit. Start with billings, deduct equipment/manufacturing acquisition, local transport, import and duties, on-site labor, subcontractor, insurance, inspection, warranty reserve, performance penalties and billing leakage. Then account for customer payment delay, advance payments, retention, VAT and financing costs. Contract margins should be computed on consistent currency and recognition bases; mixing EBITDA and gross contribution will misstate ROI.

A US$2m bid with 28% contribution before financing might still fail after large change-order disputes and long retention. Sensitivity should cover 15–35% contribution, 5–25% success probability, 90–270-day collection and 0–10% potential liquidated damages; use source-backed ranges when available, otherwise visibly marked assumed stress cases.

08 — Supplier mapping without implying endorsement

The screening output should name which commercial roles might buy, specify known procurement channels and distinguish sponsor, EPC, engineering consultant, system integrator, tier-one supplier, subcontractor and end operator. No company relationship should be implied by researching its publicly announced projects. Major-asset scale and headline investment are only rough proxies for indirect equipment demand; never convert a US$6bn data-center announcement into an alleged addressable supplier funnel without procurement packages and known contracting entities.

Specific vendor leads need dated evidence of eligibility and actual buying demand. Equipment or design scope can change between predevelopment and award. Proposed economic benefit can be estimated, but positive expected value requires both commercial access and operational delivery capability.

09 — Negative-case challenge to the thesis

Test whether the customer has no tender, the commercial timetable is wrong, technical certification is missing, the incumbent has a non-compete or exclusivity, or the supplier cannot provide financing/security. Test whether research reveals an inferior bid opportunity rather than improving conversion. Evaluate whether a low probability was rational because the procurement is already effectively closed. If qualification raises forecast win rates only by filtering out impossible deals, track both avoided cost and opportunity replacement economics, not a fictitious causal win-rate gain.

10 — What a first-class deliverable contains

An executive-grade Opportunity Qualification Packet should include: source-date register; country/regulatory summary; named commercial actor map with role and evidence status; current project stage; technical needs; procurement method; budget evidence; alternative suppliers; import/logistics dependencies; tender chronology; probability calibration; contribution and working-capital model; bid/no-bid recommendation; rebuttal and verification requests. It should include a one-page decision ledger recording what changed between the baseline and recommended allocation of commercial resources.

For monitoring engagements, additionally provide a weekly evidence/change log with significance scoring, alerts for new tenders or budget reallocations, source confidence, expiry time, and explicit separation of confirmed announcements from unresolved hypotheses.

11 — Measured value attribution

To claim verified client savings or incremental profit after implementation, establish an agreed pre-engagement baseline, log recommended actions and decision dates, capture actual spend and award outcomes, remove external confounders where possible, and measure realized contribution on collected contracts. Distinguish a bid canceled for economic reasons from a client that would have declined anyway. Quantify incremental value net of research fees and ensure that the finance team can reproduce every number from ledger entries. Without this, the US$46k example remains a useful mathematical illustration—not a case history.

12 — Falsification and compliance

Do not recommend bid pursuit when available evidence does not establish legal participation, sanctions/AML compatibility, clear procurement procedures, feasible technical delivery or a credible payment source. Avoid promising introductions, access to undisclosed documents or influence over awards. Stop if the risk-adjusted expected contribution remains negative or if an alternative market/client produces higher return on scarce business-development capacity.

13 — Evidence reference

[S1] Proyectos México official-sector electricity profile, incorporating reported 2025 system metrics. The numerical expected-value and procurement examples use explicitly disclosed scenarios; they are not official sector return statistics or historical results at named suppliers.

14 — Advanced annex: probability calibration and scarce capacity

The value of improved research depends on how much it improves causal decision quality, not on the optimism of a sales forecast. Build stage-by-stage conversion rates from dated CRM cohorts; compute Bayesian or otherwise calibrated posteriors for sparse segments, show credible intervals, and distinguish sample selection from intervention effect. If the bid probability is 8% after careful qualification, the additional US$10k spent has negative expected incremental value if no comparable uplift occurs.

With limited sales and engineering hours, optimize a portfolio of bids subject to explicit resource and risk constraints. A package with better expected dollars may use so many engineering hours that it crowds out two higher-return proposals. For every accepted opportunity, publish a traceable 'why now / why buyer / why our supplier / why economics' decision; disqualify deals on legal, compliance or operational grounds before modeling upside.

15 — Measurement design for a credible commercial result

A client should agree before implementation on a measurement protocol: time-stamped prospect universe, consistent lead definitions, pre/post bid-cost accounting, matched cohorts where possible, award and revenue recognition rules, cash-collection evidence, and treatment of pre-existing pipeline. An observed reduction in pursuit spending can be described as avoided expense only if that spend would actually have occurred absent the intervention. Report incremental after-fee profit only when contract collections, delivery costs, losses and taxes are measurable.

16 — Three-case assumption protocol

The three sets below are boundary tests, not a fitted probability distribution. Use the central reference as the working case and inspect downside/upside; never average scenario outputs without defensible weights.

Formula: ΔEV = contract value × scenario contribution margin × (qualified win probability − 8% reference win chance) − (qualified pursuit expense − US$15,000 baseline). Probabilities are uncalibrated assumptions until historical comparable win/loss outcomes are observed.

Decision rule: report the CENTRAL scenario as the provisional working case and the full range separately. Do not take the mean of three hand-picked cases; use a probability-weighted expectation only when probabilities have an auditable empirical basis.