The 0.1% Problem: Why Small Decision Improvements Matter on $100M Infrastructure Projects

The economic case for better infrastructure decisions does not require a dramatic percentage improvement: on large capital commitments, small changes in exposure, timing, or scenario selection can materially exceed the cost of decision intelligence.

Why this matters

A $50,000 decision-intelligence engagement can sound substantial in isolation. Against a $100 million capital decision, it equals 0.05 percent of project value. That arithmetic does not prove a return. It defines the economic hurdle: the engagement needs to improve the quality, timing, or conditions of the decision by only a small fraction of the capital under consideration to become material.

The responsible case is not that every identified risk becomes a saving. Exposure is not the same as loss, and an estimate is not the same as cash protected. The useful question is whether leadership has connected the project evidence to the capital, delay, financing, service, and strategic consequences while the decision is still reversible.

Where the project evidence fractures

Economic consequences often sit in separate models. The schedule shows a delay without its carrying cost. The capital model includes escalation without identifying which unresolved dependency could trigger it. Procurement tracks supplier risk while the investment paper assumes the target service date remains unchanged. Each team can be internally consistent while leadership receives no consolidated view of decision exposure.

A credible Decision Value Review must prevent false precision and double counting. It should show the source of each assumption, the period and project path to which it applies, a reasonable range, the degree of confidence, and any overlap with another exposure. Scenario-based estimates should remain distinguishable from realized savings.

• Capital under evaluation and the next irreversible commitment
• Delay, escalation, carrying-cost, financing, and service-date assumptions
• Scenario-specific exposure without double counting
• The source, owner, range, and confidence behind every estimate

What different stakeholders see

The sponsor wants momentum. Finance wants a value case. Technical leaders resist numbers that imply more certainty than the evidence supports. The capital committee wants comparable alternatives. The right response is not a theatrical ROI multiple; it is an assumption-led exposure model that shows what may be at stake, what can still be addressed, and what evidence would change the range.

A practical decision framework

The framework below is deliberately compact. It does not replace engineering, legal, financial, environmental, or regulatory diligence. It helps leadership connect those disciplines to the specific authorization being requested.

1. 1
   Define the decision and capital at stake. Record the evidence owner, current source, unresolved condition, and consequence for the decision. Where views conflict, preserve the disagreement rather than averaging it away.
2. 2
   Separate exposure from guaranteed savings. Record the evidence owner, current source, unresolved condition, and consequence for the decision. Where views conflict, preserve the disagreement rather than averaging it away.
3. 3
   Quantify a small set of material scenarios. Record the evidence owner, current source, unresolved condition, and consequence for the decision. Where views conflict, preserve the disagreement rather than averaging it away.
4. 4
   Show ranges, assumptions, and potential overlap. Record the evidence owner, current source, unresolved condition, and consequence for the decision. Where views conflict, preserve the disagreement rather than averaging it away.
5. 5
   Compare addressable exposure with the cost of the decision process. Record the evidence owner, current source, unresolved condition, and consequence for the decision. Where views conflict, preserve the disagreement rather than averaging it away.

Conditional approval is often more responsible than a false binary. Leadership can authorize limited, reversible work while withholding an irreversible commitment until named conditions are satisfied. The conditions should be observable, owned, dated, and connected to a clear decision right.

Questions leadership should ask

On large projects, decision value can become meaningful long before it reaches one percent of project cost. The strongest business case therefore connects a modest set of material exposures to current evidence, tests what is addressable before commitment, and records what leadership changed or conditioned. That is a more defensible economic story than promising savings that no prototype—or responsible advisor—can guarantee.

How the BEAVR prototype would structure this decision

In a live, human-reviewed engagement, the current BEAVR prototype would organize supplied project documents around the decision, identify evidence that is known, missing, stale, or conflicting, compare a focused set of scenarios, and produce an explainable executive decision brief. BEAVR Confidence™ communicates the strength of the connected evidence for the decision; it is not a promise of certainty or a prediction of project performance.

This is an illustrative prototype application, not a claim of a completed customer deployment, automated industry integration, or future-module availability. Specialist judgment and owner accountability remain essential.

Sources and further reading

• U.S. Government Accountability Office, Cost Estimating and Assessment Guide: Best Practices for Developing and Managing Program Costs, March 2020
• U.S. Government Accountability Office, Schedule Assessment Guide: Best Practices for Project Schedules, December 2015
• Infrastructure and Projects Authority, Project Routemap, 2021
• BEAVR Intelligence editorial standards and disclosure policy