The Identity Problem in Measurement

Why Leaders Measure What Protects Them Instead of What Reveals Them

Measuring What Matters Series

Issue 268, June 11, 2026

A division vice president I worked with at a client several years ago sat across from me with a single page in his hand. It contained five proposed additions to his quarterly scorecard. Four of them he approved readily and without comment. The fifth, a measure of how often projects under his division were re-scoped or quietly absorbed into other budgets after underperforming, he set aside with a polite but firm explanation that the methodology used to measure that proposed metric was not yet mature enough to include. The methodology was adopted by the organization and had been used in two peer divisions for more than a year. The data was clean and accurate. The dashboard already reflected the measure as staff felt it was an organizational requirement, but he once again demanded its removal. What wasn’t yet mature enough was his own willingness to see what that metric really showed and what that might say about his leadership and his positional competency.

Six months later, the same indicator measuring his division’s projects surfaced again through a divisional working group formed to address issues with a variety of failing and stalled projects. He raised the same objection he had in the prior instance. The metric was eventually removed by staff as they felt they needed to meet their leader’s criticism that it would create confusion if introduced mid-cycle.

To this day, the division vice president describes his scorecard as one of the most disciplined in the organization. He isn’t lying. He simply has no metric that contradicts him and the claim.

This is the identity problem in measurement, and it is one of the most consequential and least discussed dynamics in organizational performance. Leaders don’t just measure what’s convenient, as I have discussed in earlier pieces in the Measuring What Matters series. They measure what protects their professional identity. The metrics they elevate, the dashboards and scorecards they tolerate, the indicators they kill in committee meetings, all of these are shaped by an unconscious filter that prioritizes the preservation of how a leader sees themself and how they wish to be seen and evaluated by others. The KPI conversation is rarely just about one or more strategic priorities. It is about identity defense conducted with the language of analytical forensic-like rigor. And the cost of that defense, measured across an organization and a career, is incalculable.

The Identity Lens

Every leader carries what can be called the Identity Lens, a largely unconscious set of preferences that determines which categories of information feel safe to surface and which feel threatening. The Identity Lens is not a character flaw. It is a structural feature of being a senior executive whose professional credibility is tied to certain narratives about competence, judgment, and contribution. A CFO whose identity is anchored in cost discipline will reflexively elevate cost metrics and resist measures that would surface the long-term capability challenges produced by continuously aggressive cost cutting. A Chief Marketing Officer whose identity is anchored in brand strength will gravitate toward brand health metrics and treat with skepticism any measure that ties marketing investment directly or even indirectly to revenue. A Chief People Officer whose identity is built on culture will defend engagement scores and challenge the validity of behavioral indicators that reveal compliance without conviction. None of these leaders is acting in bad faith. They are acting in identity-protective faith, which is harder to detect because it operates inside a coherent professional logic that the leader has internalized over the years.

Chris Argyris, a researcher I often mention in this newsletter and on the podcast, spent his career documenting this pattern. His 1990 book Overcoming Organizational Defenses described what he called defensive routines, the structured ways in which individuals and organizations protect themselves from the embarrassment or threat that learning would entail. Argyris observed that the most intelligent and accomplished professionals were often the most skilled at defensive reasoning because their identity was built on being right and being seen as competent. When information threatened that identity, the response was not to integrate the information. It was to discredit it through a sophisticated process Argyris called skilled incompetence. The KPI design process is one of the purest areas where skilled incompetence operates.

Methodologies are questioned, data quality is doubted, scope is challenged, all in language that sounds like analytical rigor and is actually identity defense conducted at a high level of abstraction.

George Akerlof and Rachel Kranton extended this insight into economics in their 2010 book Identity Economics, where they demonstrated that people often make economic decisions not to maximize utility in the conventional sense but to align their behavior with the identity category they associate with. A leader whose identity category is “the disciplined operator” will systematically prefer metrics that reinforce discipline as a strategic virtue. A leader whose identity category is “the visionary” will resist any measure that requires them to be accountable for execution details. Perhaps you have experienced this or other categorization as you sat in meetings? I have seen this play out far too often over the years via engagements with one client or another.

The measurement system, then, far from being a neutral display of performance management, becomes an expression of the identities of the people who designed it. And because those leaders typically sign off on each other’s scorecards, within an organization, through some form of executive or peer review, the identity protections of one function become embedded in the dashboard architecture of the whole organization. As a result, organizations, without even realizing it, begin to measure not what matters, but what protects the identities of the people who designed the system.

Three Patterns of Identity-Driven Measurement

The first pattern is the protective omission. This is the metric that is never added because the leader who would be most accountable for it can’t tolerate or accept the reality of what it would show. The protective omission rarely manifests as an explicit refusal of inclusion. It appears as a series of small objections about methodology, timing, scope, or data quality. Because the objections are coming from a leader and most people around the table trust, or at least believe, that the goal is to support the leader, each objection starts to sound reasonable on its own. Who are those around the table to question when the leader is the one with the ultimate responsibility?

As such, the objections ensure that the inconvenient but important metric, reflecting the truth, never reaches the division, department, or organization’s dashboard. The protective omission is often visible only when you map what an organization measures against what it claims to value. The gap between the two is where the omissions live. An organization that claims customer trust is its highest priority but doesn’t measure trust beyond an annual NPS score is telling you something about the identities of the people who decide what gets measured.

The second pattern is the inflated adjacency. This is the metric that gets disproportionate attention because it makes a particular function or leader look good, regardless of whether it actually drives organizational outcomes. The inflated adjacency is identifiable by the gap between how much weight a metric receives in the dashboard hierarchy and how strongly it correlates with the strategic results the organization claims to care about. These are the metrics that occupy disproportionate visual real estate on the dashboard, rendered in larger frames, stronger colors, or heavier emphasis than the outcomes they are meant to support.

The inflated adjacency can manifest as the number of press mentions, training hours delivered, dashboards built, pipeline contacts created, and so on. Inflated adjacencies thrive in measurement systems where the leaders responsible for the function have meaningful influence over what counts as success. They wither in measurement systems that are designed by people whose identity is not tied to the function being measured.

The third pattern is what I call strategic vagueness. This is the deliberate or semi-conscious imprecision applied to the metrics that would be most diagnostic if measured cleanly. Strategic vagueness shows up in language like “directional indicators,” “qualitative measures,” and “leading proxies” applied to measurements that could in fact be defined and tracked rigorously.

The vagueness is not a methodological limitation. The data often exists to tell a story without vagueness. It is, then, a structural feature designed to preserve interpretive flexibility for the leader whose performance the metric would otherwise constrain. Sometimes, and more often than not, it is about the stories we want to tell and seeking and manipulating measurements to provide a somewhat factual basis to the story.

A culture score that aggregates twelve different survey items into a single composite, with the weights determined by the function being measured, is a strategic vagueness in operation. A digital transformation success metric defined as “progress against milestones, qualitatively assessed” is another. The vagueness ensures that no matter what happens, the leader can credibly claim that the indicator is moving in the right direction, or at least not moving in a way that requires hard accountability.

The combined effect of these three patterns is a measurement system that has the appearance of rigor and the function of protection. The dashboards are full and tell a story in what and how they represent. The reviews are disciplined. The reports are circulated on schedule. And the metrics that would reveal where leadership identity is preventing learning are systematically absent, distorted, or rendered too vague to interpret. This is not a failure of any individual or organizational analytical capability. It is the success of identity defense operating through the analytical capability of the organization.

The Functional Identity Trap

Every executive function carries an associated identity, and that identity shapes the measurement preferences of the people inside it. Finance professionals are trained to elevate precision, conservatism, and verifiable accuracy. They are uncomfortable with metrics that resist clean quantification because the discomfort threatens the core of what their professional training tells them rigorous work looks like. The result is a systematic underinvestment in measuring the intangibles that drive financial outcomes, a pattern I examined at length in The Value of Intangibles in this series. The finance function is not opposed to measuring trust, culture, or institutional knowledge. It is opposed to measuring them in ways that finance professionals consider rigorous, which means it is opposed to measuring them at all.

Sales leaders carry a different functional identity, anchored in the accumulation of revenue. The measurement preferences that flow from this identity are biased toward leading indicators of revenue (pipeline value, conversion rates, deal velocity) and resistance to indicators that would surface the long-term cost of short-term revenue tactics. Wells Fargo’s cross-selling scandal, documented extensively in the years following 2016, was in part a measurement failure produced by a sales identity that prioritized account openings and downplayed any indicator that would have revealed the unsustainable methods used to produce them. The metric that mattered most, the relationship between account openings and customer dissatisfaction over time, was not measured because no one whose identity was at stake wanted it to be. If you want to learn more about Wells Fargo, explore episode 12|025 of the Human Factor Podcast with my guest Eric Hoplin, who spent years inside Wells Fargo helping them survive their crisis.

Marketing functions tend to gravitate toward attention metrics, brand metrics, and qualitative indicators of cultural relevance. The functional identity is anchored in creativity, narrative power, and audience resonance. Marketers often resist tightly attributed revenue metrics not because attribution is impossible but because attribution would compress the role of marketing into a function whose value is tied directly to short-term outcomes, which threatens the identity that draws people into the marketing profession in the first place.

Overall, operational functions carry an identity built on efficiency, throughput, and process discipline. The measurement preferences are biased towards measures of speed, cost per unit, and adherence to the standards of work, with resistance to measures that would surface the human cost of operational discipline, including burnout, turnover among high performers, and the suppression of dissent during change.

None of these functional identities is wrong on its own. Each represents a legitimate professional perspective grounded in real expertise. The problem is not the existence of functional identities. The problem is the absence of any structural mechanism that surfaces what each functional identity prefers to see or not to see. Without that mechanism, the organization’s measurement system becomes a composite of identity protections rather than a strategic instrument for revealing reality.

Designing for Identity Honesty

The first principle of designing for identity honesty is that the leader most accountable for an outcome should not be the leader who determines how that outcome is measured. This is a structural change, not a cultural one, and it is uncomfortable for the executive teams that have spent years building scorecards in which each function defines its own success criteria. Separating accountability from measurement design is not a vote of no-confidence in any individual leader. It’s an acknowledgment that the Identity Lens operates universally, and that designing measurement systems with that knowledge in mind is the only way to surface what identity defense systematically obscures.

The second principle is paired ownership of high-stakes metrics. For any measure that has significant identity implications for a particular function, that function should share ownership with at least one peer function whose identity creates a different set of biases. An example: a measure of marketing’s contribution to revenue could be co-owned by marketing and finance, because the identity defenses in each function pull in different directions and create something closer to honest measurement when held in tension. Another example: a measure of operational efficiency should be co-owned by operations and human resources, because operations will defend efficiency metrics that suppress measures of burnout, and human resources will defend culture metrics that ignore operational impact. Paired ownership doesn’t eliminate deeply human identity defenses. But it does surface the default behavior through the friction between two professional identities that view the same situation through very different lenses.

The third principle is the mandatory uncomfortable metric. Every senior leader should have one indicator on their dashboard that they would not have chosen, designed by someone whose identity isn’t invested in the outcome the leader is responsible for. This is the metric that, when it moves in the wrong direction, the leader can’t dismiss it with the methodological objections that would have killed it in a leadership meeting, board meeting, a workgroup, or a committee. It’s the metric that, by its presence on the dashboard, signals that the leader has accepted the discipline of being measured by something they don’t control. The mandatory uncomfortable metric is the operational instantiation of what Argyris called double-loop learning, the willingness to examine the assumptions and identity structures that produced the system, not just the actions within the system. Argyris shared that organizations that build this practice into their executive review processes describe the early experience as deeply uncomfortable and the later experience as transformative. Once leaders see what they were systematically not measuring, they can’t return to the dashboards that protected them.

The fourth principle is the periodic identity audit, applied to the measurement system itself. At least once a year, an organization serious about measuring what matters should examine its dashboard architecture and ask three questions. What does this measurement system protect? Whose identity does it preserve? And what would a different set of identities have chosen to measure instead?

The questions may seem hard, and answers elusive. But they are necessary. They are uncomfortable questions because the honest answers reveal the human behavioral structure beneath the analytical surface. Without some form of an identity audit, every improvement to a measurement system risks reinforcing the very identity structures that produced its blind spots in the first place. At a time of rapid technological advancement, instability in the labor force, and global market dynamics shifting by the day, blind spots do not serve any organization well.

The Cost of Identity-Protected Measurement

The cost of identity-protected measurement isn’t abstract. It’s visible in every change or transformation effort that stalls or fails despite a dashboard showing 82 percent employee support, every product launch that surprises leadership because the warning signs weren’t measured, every strategic pivot that arrives too late because the metrics that would have revealed the need for change were never built. Sydney Finkelstein’s 2003 analysis of corporate failures, Why Smart Executives Fail, documented case after case where senior leaders had access to all the information they needed but had constructed measurement systems that filtered out the information their identities couldn’t absorb. The failures Finkelstein documented were not failures of analytical capability. They were failures of identity flexibility, expressed through the vehicle of the dashboard.

Daniel Kahneman’s work on motivated reasoning, summarized most accessibly in Thinking, Fast and Slow (2011), demonstrates that the human mind processes information not as a neutral evaluator but as an advocate for conclusions that align with existing beliefs and identity commitments. Each one of us does this, but rarely are we consciously aware that we do.

The implication for measurement design is, without question, direct. A measurement system designed by people whose identities are at stake will reliably under-represent the categories of information that would challenge those identities. It’s a very predictable behavioral output of human cognition operating under organizational pressure. The only counter to it is structural, not psychological. We can’t ask leaders to set aside their identities. We can design measurement systems that compensate for the predictable distortions that those identities introduce.

The deeper question, and the one I think every executive team should hold honestly, is not whether their measurement system is rigorous. It is whether their measurement system is honest about what it can’t see. What is being hidden? What is being overlooked? What was removed or omitted?

An organization with a less precise but more identity-aware measurement system will outperform an organization with a more precise but identity-protected one over any meaningful time horizon, because the identity-aware system will surface problems while they are still solvable. The identity-protected system will only surface them when the damage has become impossible to obscure, which is to say, when it is far too late to do anything about them.

We have spent decades treating measurement as a technical discipline. The next decade will require us to treat it as a human one. The metrics that an organization chooses to elevate, the dashboards it builds with care, the indicators it kills in committee, all of these are artifacts of the identities of the people who designed them.

There is a version of this future in which AI produces the dashboards. Algorithms that surface patterns without deference to the professional identity of the person requesting the report. Systems that measure what the data reveals rather than what the executive committee is comfortable seeing. That possibility is real, and it’s closer than most organizations appreciate. But even AI-generated measurement will be shaped by the questions it is asked to answer, the data it is given access to, and the outputs that leadership chooses to act on. The Identity Lens does not disappear when the dashboard is automated. It migrates upstream, into the design of the system, the selection of inputs, and the interpretation of results. The technology changes. The human dynamic doesn’t, unless we choose to design for it.

Until we recognize our very human defaults, every methodological improvement we make to organizational measurement will be a more sophisticated version of the same problem. The challenge isn’t to build better dashboards. It’s to build dashboards designed by people honest enough to admit what they don’t want to see.

Further Reading

The themes explored in this article connect to several other pieces in the Measuring What Matters series and the broader Ideas and Innovations archive. Each of the following offers deeper context on a specific dimension of the identity problem in organizational measurement.

The Value of Intangibles (Measuring What Matters Series) examines why 92 percent of organizational value resides in intangible assets while most measurement systems track only the tangible 8 percent, with implications for the identity defenses that keep intangibles unmeasured.

Why Transformation Dashboards Lie (Measuring What Matters Series) explores how confirmation architecture shapes measurement systems to reflect what leadership wants to see, with direct overlap on the patterns described in this piece.

The Organizational Memory Problem (Measuring What Matters Series) documents how organizations forget what they learn, including the way leadership identity protects against learning that would require an admission of past misjudgment.

The Leading Indicators of Resistance (Measuring What Matters Series) describes the behavioral signals that predict transformation failure before performance data catches up, and the identity-driven resistance to acknowledging those signals.

How Illusion and Delusion Derail Organizations (Issue 56) examines how confirmation bias and identity protection distort what organizations believe they know versus what they actually know.

Leading With Courage (Newsletter) explores why measuring what matters requires the courage to surface information that threatens existing identity structures inside leadership teams.

Sources

1. Argyris, C. (1990). Overcoming Organizational Defenses: Facilitating Organizational Learning. Allyn & Bacon.
2. Argyris, C. (1991). “Teaching Smart People How to Learn.” Harvard Business Review, 69(3), 99–109.
3. Akerlof, G.A. & Kranton, R.E. (2010). Identity Economics: How Our Identities Shape Our Work, Wages, and Well-Being. Princeton University Press.
4. Kahneman, D. (2011). Thinking, Fast and Slow. Farrar, Straus and Giroux.
5. Finkelstein, S. (2003). Why Smart Executives Fail: And What You Can Learn from Their Mistakes. Portfolio.
6. Edmondson, A.C. (2018). The Fearless Organization: Creating Psychological Safety in the Workplace for Learning, Innovation, and Growth. Wiley.