Imagine a quarterly business review where the leadership team studies the dashboard, confirms that pipeline is up, retention is stable, and feature velocity is on track. Everyone leaves the meeting feeling confident. Six months later, the strategic initiative those metrics were supposed to support has quietly stalled — and nobody can pinpoint when the disconnect began.
This scenario plays out across thousands of organizations every quarter. Research by Michael Mankins and Richard Steele, published in Harvard Business Review, found that companies on average realize only about 63% of their strategies' financial potential — with the gap attributable largely to breakdowns in planning and execution. One of the most persistent of those breakdowns is a measurement system that cannot detect when strategy and execution have parted ways. At a global scale, this means billions of dollars in strategic investment flowing through organizations whose metrics are structurally incapable of telling leadership whether the strategy is working until long after the window to course-correct has closed.
The measurements we inherit
In the organizations I have worked with, KPIs are rarely designed from scratch. They tend to be inherited. A finance team once needed revenue by region. An operations lead once cared about utilization rates. A founder once tracked pipeline because pipeline was the only metric that felt tangible in the early days.
These metrics stuck. They became standard, got added to dashboards, and eventually became what "measuring performance" means inside that company. There is nothing inherently wrong with any individual metric. The challenge lies in the absence of a deliberate process for connecting metrics across organizational levels and reconnecting them over time as strategy evolves.
It is worth drawing a distinction here that often gets blurred. KPIs, in the way I use the term, are relatively static metrics that measure the health of a business or service — revenue, system uptime, customer retention, operational throughput. These metrics reflect the ongoing state of the organization. They tend to persist regardless of which strategy is currently in play, because they measure the vision or the business model rather than the method of reaching it.
OKRs operate differently. An Objective is a transformative goal — something the organization is trying to change or achieve — and it carries its own metric for measuring progress. Each Key Result beneath that Objective also has a metric. Both levels have metrics, but the metrics are inherently tied to a specific strategic direction. When strategy shifts, OKRs should shift with it.
This distinction matters because it nuances a question I have posed before: "how many of your KPIs would change if your strategy changed tomorrow?" The honest answer might be "not many" — and that could be perfectly reasonable if those KPIs are measuring your vision rather than your strategy. The more revealing question is whether there is an explicit, testable link between your KPIs, your Objectives, your Key Results, and the deliverables your teams are actually working on. That chain of metrics — strategy metric to Objective metric to Key Result metric to deliverable metric — is where most organizations lose the thread.
Strategy is a hypothesis
Here is the reframe that can shift the entire conversation: a strategic plan is a statement of what you believe will happen, given certain conditions and certain actions. It is a bet — grounded in analysis and judgment, but a bet nonetheless.
When you say "we will grow enterprise revenue by 30% by deepening relationships with existing accounts," you are making a claim about causality. You are asserting that a specific set of actions will produce a specific outcome. That causal claim can turn out to be right, or it can turn out to be wrong. The only way to know is to test the claim, and to design your measurement system around that testing.
This is what Causal Measurement, or Kausaalinen mittaaminen, means in practice: building a measurement system that treats strategic plans as hypotheses and actively examines whether those hypotheses hold up. Confirmation-oriented measurement asks "are we on track?" Hypothesis-testing measurement asks "is the logic of our strategy actually correct?" These are quite different questions, and they lead to very different conversations in the boardroom.
Lagging indicators and the limits of hindsight
The dominant model of business measurement is built largely on lagging indicators. Revenue, bookings, retention, NPS, market share — these tell you what happened. They are essential for understanding results, and strong outcomes are precisely what organizations should be striving toward. The limitation is not that lagging indicators are unimportant but that they are insufficient for strategic management. By the time a lagging indicator moves, the window to intervene has often already narrowed or closed.
A drop in retention tells you something went wrong six or twelve months ago. A decline in NPS tells you customers experienced something disappointing — but the score alone does not reveal which assumption in your strategy failed. Organizations that rely primarily on lagging indicators often find themselves in a perpetually reactive posture: leadership gathers after the quarter closes, studies what happened, and makes adjustments, only to repeat a similar cycle of surprise. The dashboard is capturing the consequences of last year's strategy, rather than testing the validity of this year's.
Charles Goodhart captured a related dynamic in what became known as Goodhart's Law: when a measure becomes a target, it ceases to be a good measure. The principle applies directly to metrics that were designed to track one strategic hypothesis but gradually become self-justifying targets, decoupled from the strategy they were meant to test.
Vanity metrics can mask strategic failure
There is a distinct problem that often coexists with the lagging indicator challenge: measuring activity as a proxy for strategic impact.
Sales teams measure calls made, emails sent, meetings booked. Marketing teams measure content published, events attended, MQLs generated. Product teams measure features shipped, tickets closed, velocity points burned. These are real numbers, straightforward to collect — and they can create a misleading sense of accountability without revealing whether the activity is moving the organization toward its strategic goals. Eric Ries, in the Lean Startup methodology, called these "vanity metrics": numbers that look good on a dashboard but do not help you make better decisions.
High activity can coexist with minimal strategic progress. Consider a sales team making thousands of calls while the strategic thesis — that outbound prospecting to mid-market manufacturing accounts will unlock a new growth segment — is quietly failing. The call volume goes up, the activity dashboard looks healthy, and the underlying thesis goes unexamined until the end of the year, when someone notices that the new segment never materialized.
Causal Measurement pushes toward a harder question: rather than "did we do the activities we planned?", the question becomes "did those activities produce the intermediate outcomes we expected?" Vanity metrics sit at the beginning of the causal chain. Meaningful impact metrics sit further along. In my experience, organizations often measure the activity and assume the strategic impact follows — which sometimes it does, but not reliably enough to build strategy on.
The consequence of these two patterns — lagging indicators and vanity metrics — is worth naming directly. A measurement system that cannot tell you when your strategy is struggling risks becoming a form of progress theater, however well-intentioned the people running it may be. Across an entire economy, this structural weakness means that organizations are collectively making strategic investment decisions based on metrics that cannot distinguish between genuine progress and comfortable momentum.
A diagnostic for the curious practitioner
Here is a practical exercise that takes less than fifteen minutes and often reveals more about your measurement system than any dashboard review.
Pick your top three KPIs — the ones that appear in your board materials and executive scorecards. For each KPI, try to draw the full chain of metrics: from your strategic intent, to the Objective it connects to, to the Key Results beneath that Objective, to the deliverables your teams are working on. Not a loose association — an explicit chain where each level has its own metric and the link between levels is testable.
Then ask three questions about each chain:
Is the chain explicit? Was this metric linkage articulated before you started executing, or are you constructing the logic now to explain a number that is already moving?
Is the chain testable? Are there intermediate metrics in the chain that you are actually measuring — checkpoints that could tell you early whether the causal logic holds up in practice?
Is the chain still current? Does this chain reflect your strategy today, or does it reflect a strategy you had twelve, eighteen, twenty-four months ago?
If you cannot draw the chain for a KPI, that KPI may be measuring something other than your current strategy. If the chain exists but lacks testable checkpoints, you are hoping rather than testing. If the chain is outdated, you are measuring yesterday's strategy with today's resources.
In my experience, practitioners who run this exercise find that one or two of their top KPIs survive the scrutiny well. The rest tend to reveal a measurement system that drifted from strategic intent — layer by layer, quarter by quarter — until the chain of metrics became difficult to trace. That drift is a design characteristic, not a personal failure. And it is one that can be corrected by anyone willing to ask what their metrics are actually measuring.