Three Structural Repairs That Stop the Bleeding From Broken Performance Measurement

The diagnosis is in. Manager engagement at 22%, half of all projects failing to deliver value exceeding their cost, trust in immediate managers down 37% in three years, and KPIs being gamed at scale. You already know something is wrong. The question is what you actually do about it this week, not what framework you admire next quarter. These three repairs are sequenced, specific, and runnable. Each one addresses a structural cause, not a symptom.

Repair 1: Audit Your Top Five Metrics for Goodhart's Law Vulnerability

Goodhart's Law is simple: when a measure becomes a target, it ceases to be a good measure. This is not a theory about bad people. It is a description of how optimization pressure works. When you set a number as the goal, you create incentive to hit the number through the path of least resistance, which is often not the path that produces the underlying outcome you wanted. The performance review never catches this because the number looks fine.

Here is how to run this audit. Pull your top five performance metrics — the ones that appear in executive reviews, that managers are accountable for, that drive compensation decisions. For each metric, work through these four questions:

When you find gaming signatures, you have two options. The first is to rotate the metric. This works short-term but does not fix the underlying problem because people will optimize against the new number too. The second option is what actually works: pair your primary metric with a second-order measure. Pairs of metrics are significantly harder to game simultaneously than individual metrics. If your primary metric is call volume, your second-order measure might be first-call resolution rate. Hitting both requires producing the actual outcome. You cannot inflate one without deflating the other.

Run this audit in one working session. Bring whoever owns each metric into the room. The goal is not to call anyone out. The goal is to identify where your measurement system has drifted from the behavior it was built to observe. For each metric where you identify a gaming signature, assign a second-order measure and set a 30-day review date to confirm the pairing is working.

One more thing to check: are you measuring strategic alignment anywhere? PMI's research across 5,800-plus respondents found that high-performing organizations use 44% more metrics than average, specifically including strategic alignment and outcome measures. If your active initiative reviews only track schedule and budget, you are optimizing for delivery optics, not value delivery. Add one alignment question to every active project review: does this initiative still address the problem it was scoped to solve? That single question surfaces drift before it becomes failure.

Repair 2: Fix the Management Span Problem

Span of control is not a soft HR concept. It is an operational variable with measurable effects. When a manager's team doubles in size without any reduction in role scope, something has to give. What gives is the quality of individual attention, the speed of decision-making, the depth of coaching, and the manager's own capacity to think strategically rather than reactively. This is the structural mechanism behind the 22% manager engagement number.

Here is how to calculate whether you have a span problem. For each of your top five managers, pull three numbers: current direct report count, direct report count 36 months ago, and any changes in the manager's role scope or administrative responsibilities over the same period. If the direct report count increased more than 25% without a corresponding reduction in scope, you have a structural overload condition. The engagement survey will describe this person as disengaged. That description is wrong. Disengaged is a label for a symptom. Overloaded is the cause.

Sustainable span of control is not a fixed number. It depends on team complexity. A manager running a team of five who all do highly interdependent work in a fast-changing environment has a heavier cognitive and relational load than a manager running a team of ten who work independently on stable processes. To assess complexity, score each team on three factors: interdependence of work across team members, rate of change in goals or methods, and the degree to which each team member requires individualized development support. High scores on all three mean a manager can sustainably handle fewer direct reports before quality of management degrades.

The two-step plan to restructure is straightforward. First, identify the managers whose spans are structurally overloaded based on the audit above. Do not start with a survey or a one-on-one asking how they feel. Start by documenting what structural support was removed during the last flattening. Was a layer of team leads eliminated? Was administrative work added at the same time headcount grew? Name the specific structural changes that created the load condition.

Second, restore structural capacity before you run another training program or engagement initiative. This might mean adding a team lead role back into the structure. It might mean removing administrative responsibilities that should not sit with a people manager. It might mean reducing span by redistributing direct reports across managers whose spans are sustainable. What it cannot mean is sending an overloaded manager to a management effectiveness training and expecting the number to move. Structure produces the condition. You fix the condition by fixing the structure.

Repair 3: Rebuild Project Planning-to-Execution Alignment

The planning-to-execution disconnect is cited by 35% of project managers as the top barrier to delivering value, according to PMI's research across 5,800-plus respondents. This is not a strategy quality problem. The plan was usually fine when it was written. The failure happens at the handoff between planning and execution, where context changes, resource assumptions drift, and the original problem the project was solving quietly stops being the problem the team is working on.

The specific failure point is this: most projects have a kickoff meeting that reviews scope, budget, and timeline. Almost none have a mechanism that confirms the project is still solving the right problem at the moment execution begins. By the time you have finished planning a complex initiative, the environment has often shifted enough that the original problem definition is partially obsolete. If no one checks, execution begins against a stale problem statement.

Run this four-question review before any project moves from planning to execution, and repeat it at every major governance checkpoint:

These four questions take less than 20 minutes in a governance meeting. They surface misalignment before it becomes a failed project. Given that the baseline success rate for project value delivery is 50%, any mechanism that shifts even two or three projects per year from the failure category to the success category has a measurable return on the time invested.

How to Sequence These: Do Not Try All Three at Once

Three simultaneous structural repairs will produce zero completed structural repairs. Sequence matters. Here is the order of operations.

Start with the metric audit if your performance data is unreliable or you suspect your teams are hitting numbers without producing outcomes. Broken measurement is the most upstream failure. If you cannot trust the signal your system is producing, the other two repairs will be evaluated against bad data.

Start with the management span audit if your organization recently went through a flattening or reorganization, or if you have identified specific managers who are underperforming without a clear capability explanation. Span overload degrades everything downstream: coaching quality, decision speed, team clarity, and retention. Fixing structure at the management layer has the widest blast radius of the three repairs.

Start with the project alignment review if you have a specific initiative in-flight that is at risk, or if your portfolio review process is producing consistent surprises at delivery. This repair is the most contained and the fastest to implement. You can run the four-question review in your next governance meeting without any structural change to the organization.

The honest answer for most operators: start with the metric audit, because the other two repairs will require you to measure progress, and if your metrics are already compromised, you will not be able to tell whether the structural repairs are working. Spend the first two weeks on the audit. Move to span in week three. Layer in the project alignment questions in parallel with the span work once you have a clear baseline on the metric side.

What Success Looks Like at 90 Days

Do not evaluate structural repairs against the same lagging indicators that revealed the problem. Engagement scores, project success rates, and trust metrics all move slowly and reflect conditions from 12 to 24 months prior. You need leading indicators that tell you whether the structural changes are holding before the lagging metrics confirm it.

At 30 days, the leading indicator for the metric repair is simple: can you name the second-order measure for each of your top five primary KPIs, and is someone actively tracking both? If yes, the pairing is in place. If people are still reporting only on the primary metric, the repair has not been implemented regardless of what was agreed in the meeting.

At 60 days, the leading indicator for the span repair is manager-reported clarity on priorities. Not engagement. Not satisfaction. Ask each manager you restructured one question: compared to 60 days ago, do you have more or less clarity on what to do this week? Clarity is the proximate outcome of reduced cognitive overload. If clarity is up, the structural change is working. If clarity is flat, the scope was not actually reduced even if the headcount was redistributed.

At 90 days, the leading indicator for the project alignment repair is the number of projects that triggered a rescoping conversation before execution began rather than after the first missed milestone. If your four-question review is surfacing misalignment at the planning-to-execution boundary, the mechanism is working. If every project is passing all four questions without discussion, the questions are not being taken seriously in the review meeting.

One additional 90-day check: run your last three completed projects through the PMI value delivery test. Did each one deliver value exceeding time, cost, and effort? Not whether it shipped. Whether it delivered the underlying value it was scoped to produce. That number is your structural baseline. Anything above 50% means the repairs are working. Anything below 50% means you are still operating at the industry average for project failure, and the planning-to-execution repair needs more attention.

The Close

None of this requires a consultant, a new software platform, or a multi-quarter transformation program. The metric audit is a two-hour working session. The span calculation is an afternoon of pulling org chart data. The four-question project review is a standing agenda item. The structural problems producing your performance data are real, but they are also specific and correctable. Run the audits, name the conditions, fix the structure, and measure the leading indicators. That is the whole job.