Look at the dashboard your leadership team actually watches, the handful of numbers that get reviewed every week, and ask a simple question about each one: is this on the dashboard because it genuinely drives the business, or because it was easy to get out of the system? For an uncomfortable number of the metrics you track, the honest answer is the second one. It is there because the data was available, the number was clean, and pulling it took no effort, not because anyone decided it was one of the few things that actually determines whether the business is winning.

This is not a small problem, because what you measure is what you manage. The metrics on the dashboard shape where attention goes, what gets discussed, what gets optimized, and what gets ignored. And if the selection of those metrics was quietly driven by what was convenient to extract rather than what was important to know, then your entire management attention is being pointed by your reporting system's path of least resistance, not by your strategy. The dashboard looks like a set of deliberate choices about what matters. Often it is a record of what was easy.

The streetlight effect, applied to your dashboard

There is a well-worn parable that captures this exactly. A man is searching for his lost keys under a streetlight. A passerby asks where he dropped them, and he points into the dark and says over there, but the light is better here. Decision scientists call this the streetlight effect, and it describes the universal tendency to look where looking is easy rather than where the answer actually is.

Applied to measurement, it becomes a design principle nobody consciously chose. As one analysis puts it, organizations end up measuring what is easy, and in doing so they don't just notice those things, they make them important. The metric that is cheap to extract gets tracked, tracking it draws attention, attention makes it feel important, and its importance justifies continuing to track it. Meanwhile the thing that actually matters but is hard to measure sits in the dark, unmeasured, and therefore undiscussed, and therefore treated as if it were less important, or as if it did not exist at all.

That last step is the dangerous one, and it has a name.

The four-step descent

The sociologist Daniel Yankelovich described the full mechanism in 1971, and it is now known as the McNamara Fallacy, after the US Defense Secretary whose data-driven management of the Vietnam War made body counts the measure of progress precisely because they were countable. Yankelovich laid it out as a four-step descent:

First, measure whatever can be easily measured. This is fine as far as it goes. Second, disregard what cannot be easily measured, or assign it an arbitrary number. Third, presume that what cannot be measured easily is not important. Fourth, conclude that what cannot be easily measured does not exist.

Each step feels reasonable from inside the previous one, which is what makes the descent so easy. You start by measuring the convenient thing, a defensible practical choice. You end, without ever making a conscious decision to, having redefined your entire sense of what matters down to the set of things your systems happen to produce cleanly. The hard-to-measure factors, which are frequently the ones that actually determine success, have been quietly legislated out of existence, not because anyone judged them unimportant but because they were never under the light.

What this looks like in a property business

The pattern is concrete once you look for it, and property has a characteristic version. The easy numbers, the ones the system produces without effort, dominate the dashboard: occupancy percentage, rent collected, work orders closed, gross rental income. These are all real and worth knowing. But notice what they share, they are all easy to count, and notice what tends to be missing, the things that are hard to count but often matter more.

Occupancy is on every dashboard because it is trivial to calculate. But the quality of that occupancy, whether you filled the building with tenants who will renew and pay reliably or with tenants who will churn and cost you, is harder to measure and usually absent, even though it may matter more to the long-run value of the asset. Work orders closed is easy; whether the underlying problem was actually resolved or will recur, and whether the tenant experience was good, is hard, and so it is rarely tracked. Rent collected this month is easy; the slow erosion of tenant satisfaction that will show up as non-renewals in eighteen months is hard, invisible on the dashboard, and building silently in the dark.

None of the easy metrics are wrong. The problem is that the dashboard, taken as a whole, systematically over-weights what is countable and under-weights what is consequential, and because leadership manages to the dashboard, the business ends up optimizing the countable at the expense of the consequential. You get a portfolio that looks excellent on every easy metric and is quietly deteriorating on the hard ones nobody is watching.

Measurement is an act of governance

The deeper point, and the one worth sitting with, is that choosing what to measure is not a neutral technical decision. It is an exercise of power over the organization's attention, and therefore over its behavior.

Because metrics shape reality rather than merely describe it, the selection of metrics is, in effect, a governance decision about what the organization will care about. When you put a number on the dashboard, you are not just observing the business, you are directing effort toward that number and away from everything not on the board. So the question of which metrics earn a place is far more consequential than it is usually treated. It is typically handled as a reporting question, decided by what the system can produce and delegated to whoever builds the dashboard, when it is actually a strategic question about where the organization's attention should go, which deserves to be decided by the people responsible for strategy.

The failure is one of ownership as much as analysis. Nobody chose to let extraction cost determine the company's priorities. It happened by default, because the easy metrics flowed onto the dashboard on their own and the hard ones required a deliberate effort that nobody was tasked with making. This is a close cousin of a pattern that recurs across systems, where a consequential structure gets set by convenience and then never revisited, examined for financial reporting in your chart of accounts is a strategy document. Here the structure is the set of things you have decided, mostly by accident, are worth knowing.

The honest part

Several qualifications keep this from becoming an argument against quantification, which would be the wrong lesson entirely.

Easy metrics are not bad metrics. Occupancy, collection, and the rest are easy to measure and genuinely useful, and the fact that a number is convenient to extract does not make it worthless. Plenty of important things are also easy to measure, and the goal is not to distrust a metric because it was easy, but to make sure the dashboard is not composed only of easy metrics while the hard, important ones go unwatched.

Quantification itself is not the enemy either. The McNamara Fallacy is sometimes misread as an argument against measuring things, and that reading is dangerous, because managing on pure intuition with no numbers is generally worse than managing on imperfect numbers. The fallacy is not measuring, it is concluding that only the measurable exists. The correct response is to measure what you can and to explicitly hold space for the important factors you cannot yet measure well, not to abandon measurement.

And some things are genuinely hard or impossible to measure, and forcing a bad number onto them can be worse than acknowledging them qualitatively. A crude, misleading metric for tenant satisfaction can do more harm than an honest statement that you are tracking it through judgment and direct contact while you work on a better measure. The goal is not to quantify everything at any cost. It is to stop letting the ease of measurement decide what counts.

The question to ask of every metric

The practical discipline is to separate two questions that get silently merged: how important is this to the business, and how easy is it to measure. These are independent, and confusing them is the whole problem, because the dashboard tends to fill up with things that score high on easy regardless of how they score on important.

Run your actual metrics through that pair. The genuinely useful ones are important and measurable, and they belong on the dashboard. The dangerous category is the reverse: the factors that are important but hard to measure, because those are the ones the streetlight effect leaves in the dark, and they are frequently where the business is actually won or lost, tenant satisfaction, staff capability, asset condition, the quality behind the quantity. For each of those, the task is not to ignore it because it resists measurement, but to find the best available proxy, accept that an imperfect measure of the right thing beats a perfect measure of the wrong thing, and put it on the board even if the number is rougher than you would like.

The single question that exposes the whole pattern: if you listed the three or four things that genuinely determine whether this business succeeds over the next five years, how many of them are actually on your dashboard, and how many are absent because they were hard to measure? The gap between that list and your dashboard is the distance between what matters and what was easy. Closing it is not a reporting task. It is deciding, on purpose, what your organization will pay attention to, which was always the real job the dashboard was quietly doing for you.

FAQs

Q1. What is the streetlight effect in the context of metrics?
It is the tendency to measure what is easy to measure rather than what matters, named after the parable of a man searching for his keys under a streetlight because the light is better there, even though he lost them elsewhere. Applied to dashboards, it means the metrics that are cheap to extract get tracked and, by being tracked, come to feel important, while the things that are hard to measure but consequential go unwatched and are treated as if they matter less.

Q2. What is the McNamara Fallacy?
It is a four-step descent described by Daniel Yankelovich: measure what is easily measured, disregard what cannot be easily measured, presume that what cannot be measured is unimportant, and finally conclude that what cannot be measured does not exist. It is named after Robert McNamara, whose reliance on countable metrics like body counts in Vietnam mistook what was measurable for what mattered. Each step feels reasonable from inside the last, which is what makes it so easy to slide down.

Q3. Are easy-to-measure metrics inherently bad?
No. Occupancy, rent collection, and similar metrics are easy to measure and genuinely useful, and convenience does not make a metric worthless. Many important things are also easy to count. The problem is not any individual easy metric but a dashboard composed entirely of them, which systematically over-weights what is countable and leaves the hard, consequential factors unwatched. The goal is balance, not distrust of convenient numbers.

Q4. What kinds of important things get left in the dark?
The factors that are hard to quantify but often decisive: the quality of your occupancy rather than just its level, whether a closed work order actually resolved the problem, tenant satisfaction that predicts renewals, asset condition, and staff capability. These resist easy measurement, so they tend to be absent from dashboards, which means the business optimizes the countable metrics while these quietly deteriorate unnoticed.

Q5. Why is choosing what to measure a governance decision?
Because metrics shape behavior, not just describe it. Putting a number on the dashboard directs organizational attention and effort toward it and away from everything not measured. So selecting metrics is really deciding what the organization will care about, which is a strategic choice. Treating it as a technical reporting task, decided by what the system can easily produce, hands a strategic decision to convenience by default.

Q6. Isn't this just an argument against using data?
No, and that reading is dangerous. Managing on pure intuition with no numbers is generally worse than managing on imperfect ones. The McNamara Fallacy is not the act of measuring, it is concluding that only the measurable exists. The right response is to measure what you can while explicitly holding space for important factors you cannot yet measure well, not to abandon quantification.

Q7. What if something genuinely can't be measured well?
Then acknowledging it qualitatively is often better than forcing a misleading number onto it. A crude, distorting metric for something like tenant satisfaction can cause more harm than honestly stating you are tracking it through judgment and direct contact while you develop a better measure. The aim is to stop letting ease of measurement decide what counts, not to quantify everything at any cost.

Q8. How do I tell if my dashboard has this problem?
Separate two questions for each metric: how important is it, and how easy is it to measure. Then list the three or four things that genuinely determine the business's success over the next five years and check how many are actually on your dashboard. The ones that are important but absent, usually because they were hard to measure, reveal the gap between what matters and what was convenient, and closing that gap is the real work.