Research / The measurement asymmetry

The measurement asymmetry

The measurement asymmetry: why we cut costs but can't see lost opportunities

2026-03-27 · cost-performance workforce-skills knowledge-management · medium · source → · wiki →
key claims
  1. Accrual-based financial reporting records only realised events, meaning forgone revenues from a cost-cutting decision are absent from the accounting system by design, not by oversight, making opportunity cost structurally invisible regardless of the quality of individual decision-making
  2. Cognitive loss aversion, established by Kahneman and Tversky's prospect theory (1979) with an estimated loss-aversion coefficient of approximately 2.25, means the psychological impact of a cost saving is felt roughly twice as intensely as an equivalent forgone opportunity gain, biasing decision-makers toward cost cuts even when the expected value of the opportunity investment is higher
  3. Present bias creates an additional discount on future opportunity returns: a 2005 survey of executives found that 78% of CFOs admitted to delaying or cancelling projects with a positive NPV in order to meet quarterly earnings targets, demonstrating that institutional present bias operates as a conscious decision, not just a subconscious tendency
  4. Attribution failure prevents the feedback loop from closing because the causal chain from a procurement decision to a lost customer conversion spans months, crosses departmental boundaries between procurement and sales, and is confounded by external market variables that no standard reporting system attributes back to the original decision
  5. Standard procurement KPI frameworks measure cost reduction, cost avoidance, and spend under management as primary objectives and contain no metric for revenue opportunity destroyed by supplier consolidation or channel changes, meaning the information architecture is designed to be blind to one side of the trade-off
  6. Sutherland's online travel company example illustrates the conversion asymmetry directly: website visitors convert at approximately 0.5% while phone callers convert at approximately 30%, a 60x difference; eliminating the phone channel saves the call-centre cost but destroys the higher-converting channel, with no line item in the business's reporting to record the loss
  7. The IPA Effectiveness Databank (covering 996 cases, 700 brands, 83 categories over 30 years) demonstrates that sales activation effects decay within weeks while brand-building effects last years; companies that shifted entirely to measurable performance marketing saw activation efficiency metrics improve while market share simultaneously declined
  8. McKinsey's 2001–2014 corporate horizons study found that long-term-oriented companies grew cumulative revenue 47% more than short-term peers, grew earnings 36% more, and produced economic profit 81% higher, a gap of sufficient magnitude to imply that the compounded effect of systematically forgoing opportunity investments is a primary driver of long-run performance divergence

Research Question

Why is measuring opportunity cost systematically harder than measuring direct costs, and what cognitive and structural mechanisms cause organisations to destroy value while claiming efficiency gains?

Findings

Executive Summary

Opportunity cost is structurally harder to measure than direct cost because accrual accounting records only realised events, and forgone revenues require a counterfactual that the accounting system does not construct. Four reinforcing mechanisms compound this: cognitive loss aversion (losses feel ~2.25x more significant than equivalent gains), present bias (78% of CFOs admit delaying positive-NPV projects to hit quarterly targets), attribution failure (the causal chain from cost-cutting decision to lost revenue spans months across departmental boundaries), and KPI structures that reward cost reduction with no countervailing revenue-opportunity metric. The result is a systematic institutional bias toward decisions that look efficient but destroy value, most visible in procurement consolidation, channel elimination, and short-cycle performance marketing over brand investment.

Key Findings

  1. Accrual-based financial reporting records only realised events, meaning forgone revenues from a cost-cutting decision are absent from the accounting system by design, not by oversight, making opportunity cost structurally invisible regardless of the quality of individual decision-making. [high confidence]

  2. Cognitive loss aversion, established by Kahneman and Tversky's prospect theory (1979) with an estimated loss-aversion coefficient of approximately 2.25, means the psychological impact of a cost saving is felt roughly twice as intensely as an equivalent forgone opportunity gain, biasing decision-makers toward cost cuts even when the expected value of the opportunity investment is higher. [medium confidence]

  3. Present bias creates an additional discount on future opportunity returns: a 2005 survey of executives found that 78% of CFOs admitted to delaying or cancelling projects with a positive NPV in order to meet quarterly earnings targets, demonstrating that institutional present bias operates as a conscious decision, not just a subconscious tendency. [high confidence]

  4. Attribution failure prevents the feedback loop from closing because the causal chain from a procurement decision to a lost customer conversion spans months, crosses departmental boundaries between procurement and sales, and is confounded by external market variables that no standard reporting system attributes back to the original decision. [high confidence]

  5. Standard procurement KPI frameworks measure cost reduction, cost avoidance, and spend under management as primary objectives and contain no metric for revenue opportunity destroyed by supplier consolidation or channel changes, meaning the information architecture is designed to be blind to one side of the trade-off. [high confidence]

  6. Sutherland's online travel company example illustrates the conversion asymmetry directly: website visitors convert at approximately 0.5% while phone callers convert at approximately 30%, a 60x difference; eliminating the phone channel saves the call-centre cost but destroys the higher-converting channel, with no line item in the business's reporting to record the loss. [medium confidence]

  7. The IPA Effectiveness Databank (covering 996 cases, 700 brands, 83 categories over 30 years) demonstrates that sales activation effects decay within weeks while brand-building effects last years; companies that shifted entirely to measurable performance marketing saw activation efficiency metrics improve while market share simultaneously declined. [high confidence]

  8. McKinsey's 2001–2014 corporate horizons study found that long-term-oriented companies grew cumulative revenue 47% more than short-term peers, grew earnings 36% more, and produced economic profit 81% higher, a gap of sufficient magnitude to imply that the compounded effect of systematically forgoing opportunity investments is a primary driver of long-run performance divergence. [high confidence]

  9. Marketing mix modelling can construct implicit counterfactuals from aggregate historical data and partially quantify opportunity cost retrospectively, but requires 12–26 weeks of pre-existing variation and its outputs become available after the cost saving has already been booked, preserving the sequencing asymmetry between visible cost and invisible opportunity. [high confidence]

  10. The most structurally tractable intervention is pre-decision baseline measurement: establishing which metrics will be monitored and what a detectable impact threshold looks like before a cost-reduction decision is executed, converting the counterfactual from retrospective speculation to prospective hypothesis testing. [medium confidence]

Assumptions

Analysis

A two-layer structure accounts for the evidence. Structurally, accrual accounting, quarterly reporting timescales, and procurement KPI frameworks are designed to measure realised costs and revenues, not opportunity cost. Measurement asymmetry is a side-effect of the system operating as intended, not a failure of individual decision-makers.

[inference] Cognitively, loss aversion and present bias ensure that even where alternative measurement approaches exist, decision-makers do not voluntarily adopt them. A finance team that could commission an MMM study before a procurement decision will not do so unless the incentive structure gives weight to the output.

[inference] Both layers interact: structural gaps make cognitive biases more consequential, and cognitive biases reduce political appetite for structural reform. [inference] Changing procurement KPIs to include a revenue-impact measure will face resistance precisely because the new metric makes previously invisible opportunity costs visible, creating accountability for decisions that previously had none.

Binet and Field's IPA Databank evidence is notable because it provides a large-scale, cross-industry empirical test rather than a case study or theoretical argument. Degradation of brand value through performance-marketing over-investment is a documented empirical pattern at scale. Companies in the IPA dataset that destroyed long-term brand value were not behaving irrationally; they were responding rationally to the measurement systems they operated within.

For AI deployment decisions, this analysis is immediately applicable. Cost savings from automating a customer service function are a known number; the opportunity cost (degraded customer lifetime value, erosion of cross-sell capture) is unmeasured with standard reporting tools. [inference] Given the structural sequencing asymmetry documented in §2, AI deployment in high-contact roles is likely to precede the establishment of an adequate opportunity cost measurement framework.

Risks, Gaps, and Uncertainties

Open Questions

sources

← newerRory Sutherland's core tenets
older →Customer contact as strategic signal

Connected items

Loading…

View full knowledge graph →