Organizations track more KPIs than ever and execute less of their strategy. Here is the data behind the KPI paradox - and the subtraction discipline that fixes it.
- Cut the plan to five to nine goals and about nine to eleven measures. Default scorecard math lands at 60 to 105 measures.
- Give every measure one real owner who updates it monthly. Objectives with an active owner are on-track 23.6% of the time, against 9.7% without one.
- Pair every gameable KPI with a counter-metric in the same packet. Time-to-Fill can hit target while recruiters reject more resumes and new-hire retention slides.
- Segment composite KPIs by revenue tier, geography or channel. A rising average satisfaction score can hide unhappy large accounts.
- Schedule a subtraction step in every annual planning cycle. Benchmark your list against the sector data in the ClearPoint Strategic Planning Report.
It is a Tuesday in December. The performance coordinator for a mid-sized city emails me at 4:47pm. Her plan carries well over a hundred measures. Twenty-eight of them are due December 31. Public Works has not updated their work-order completion rate in fourteen weeks. She is writing the third "friendly reminder" of the month. By Thursday she will be writing the department's paragraph herself. By the council meeting on Tuesday night, that paragraph will be in the packet under someone else's name.
If you have lived a version of that week, the next sentences are the ones that matter.
Across hundreds of organizations and thousands of strategic plans on the ClearPoint platform, the same structural pattern repeats: dashboards keep growing, and strategy execution does not keep up. The framework expands every planning cycle; the discipline to prune almost never ships with it. And the data points to what fixes it. The dashboards keep growing. The strategies do not.
This is what I call the KPI paradox: organizations are tracking more KPIs than ever, and executing less of their strategy. The cause is the doctrine, not technology, talent or budget.
An admission about the doctrine I trained under
I was trained on the Balanced Scorecard by the people who built it, and I have spent a career helping more than a thousand organizations implement it. I will say this plainly: the default measurement footprint that the Balanced Scorecard and most OKR rollouts recommend is several times what the data shows actually works. Four perspectives times five-to-seven objectives times three-to-five measures lands you at 60 to 105 measures. The plans that actually complete what they start carry a fraction of that. We added Learning & Growth as a perspective in the early scorecard work. We never built a counterweight to remove anything. Two decades later, the conclusion is hard to avoid: every framework deserves a subtraction discipline, and almost none of them ship one.
If that sounds like a contrarian claim against the doctrine that built this category, it is. Here is the data behind it.
How we counted
A statistic without a methodology is a hot take, so here is exactly where these numbers come from. They are drawn from the live strategic plans on the ClearPoint platform — 200+ organizations across local and state government, healthcare, higher education, utilities, financial services, technology, and nonprofits. We excluded demo and template content (the sample data that ships with the product) so the figures reflect real plans, not training sandboxes. A few working definitions, because they change what the numbers mean:
| Term | How we define it |
|---|---|
| "No active owner" | A measure or objective with no person assigned to keep it current. |
| "On-track" (green) | An item whose latest reported status is green. Of the items that carry any rating at all, roughly 56% are green; ClearPoint shows which items still need one. |
| "Completed" | An initiative flagged complete. Initiatives with a named owner are completed at 2.5× the rate of those without one (21.5% versus 8.7%). |
| "Owner effect" | The difference one field makes: initiatives with a named owner complete at 2.5× the rate of those without one (21.5% versus 8.7%). |
If you are a journalist, researcher, or buyer evaluating this analysis, the broader per-sector breakdown ships in the ClearPoint Strategic Planning Report. We publish the definitions because it is the only honest way to put statistics like these in front of strategy professionals.
What the plans actually show
The shape of the median plan is now bigger than what any framework actually recommends. The shape of the plans that complete what they start is dramatically smaller. The gap between the two is the entire paradox.
- Most plans track far more than they review. The median plan carries well over a hundred measures; the plans that finish what they start carry a fraction of that.
- Ownership is the dividing line. Assigning one roughly doubles the odds the item ends up on-track.
- Assessment is the next lever. ClearPoint shows which measures are on track and which still need a status.
- Ownership drives completion. The KPI list is not the bottleneck. The cadence and the ownership are.
Where the paradox lives — sector by sector
Regulated sectors feel the most pressure to add measures.
Two patterns stand out.
Healthcare has the most room to grow. The cause-and-effect is real and visible in our customer notes. Every regulator (CMS, the Joint Commission, state health departments, payer scorecards, accreditation bodies) adds metrics. Almost none of them subtract. The plan grows because regulatory pressure adds, and the strategy office rarely has the authority to remove. The measurement footprint balloons; the execution does not follow.
Government and education feel the most pressure to add metrics. Council and accreditation cycles add the most pressure for "just one more metric," and almost no one is tasked with retiring the old ones.
Three anti-patterns — and what each one cost
I have watched the same three KPI mistakes destroy strategic momentum again and again. The three cases below are illustrative composites: anonymized, representative scenarios drawn from patterns I have seen repeatedly, not measurements pulled from any single customer's account. They are here because the mechanism is what matters, and the mechanism is always the same.
Anti-pattern 1 — The unsegmented composite (illustrative: a mid-size electric utility)
Take an investor-owned electric utility that reports a single composite Customer Satisfaction score on the board scorecard. The number climbs over two reporting cycles. The dashboard turns green. Commercial-account churn climbs at the same time. The composite is a population-weighted average: happy small residential accounts (the bulk of the book by count) drown out unhappy large commercial accounts (a large share of the book by revenue). The fix is to split CSAT into separately reported segments by revenue tier. Within two cycles, the commercial-tier conversation shifts from "the score is fine" to a serious pricing review. The single number had been hiding a real revenue risk in plain sight.
Lesson: segment KPIs by the cohort that matters — revenue tier, geography, channel, vertical. Composite averages are where strategic blind spots hide.
Anti-pattern 2 — The activity substitute (illustrative: a regional commercial bank)
Take a regional bank that tracks "Training Hours Completed" as the headline measure for the health of its anti-money-laundering program. Completion hits 100% quarter after quarter. In the same window, regulator-flagged violations rise. Hours in chairs is not behavior change. The fix is to replace the activity KPI with a composite that pairs test-pass rate, time-to-competence, and recurrence of the same violation. That measure puts the training team and the compliance officer in the same conversation for the first time. The training-hours number is retired entirely, and the new measure surfaces the recurrence pattern within a single quarter.
Lesson: when a KPI measures activity (hours logged, tickets opened, courses completed), pair it with an outcome KPI. One without the other is theater.
Anti-pattern 3 — The Goodhart trap (illustrative: a fast-growing city HR department)
Take a city HR department that sets an aggressive Time-to-Fill target. They hit it. The dashboard turns green. What the dashboard cannot see: recruiters have quietly raised the resume-screen rejection rate to make the number. Quality of hire, measured by twelve-month retention of new hires, slides over the same window. The council never connects the two numbers because they live on different reports. The fix is to pair Time-to-Fill with twelve-month new-hire retention as a non-negotiable counter-metric, and to put both numbers in the same packet. Time-to-Fill drifts back up; retention rebuilds. The drift was the point: the dashboard had been hiding the trade-off.
Lesson: Goodhart's Law: when a measure becomes a target, it stops being a good measure. Pair every gameable KPI with a counter-metric, and put both numbers in the same conversation.
What ClearPoint killed from its own scorecard
Every framework I have recommended, I have first imposed on ClearPoint Strategy itself — including the bad versions. Here is our own kill list.
| KPI we tracked | Why we killed it | What we replaced it with |
|---|---|---|
| Monthly Active Users | Looked good, did not predict expansion. We had quarters where MAU rose and net revenue retention slipped — proof the metric was decoupled from the outcome it was supposed to track. | Active Owners per Plan — owners who updated their measure in the last 30 days, divided by total owners. It tracked our retention far more faithfully. |
| Tickets Closed per Week | Goodhart in real time. The team closed trivial tickets faster while substantive ones aged in the queue. Average resolution time looked great; churn risk did not move. | Median Time-to-First-Substantive-Reply plus % of tickets touched within 24 hours. Two specific behaviors, no gameable composite. |
| Number of Blog Posts Published | The classic activity substitute. Output volume, zero outcome signal. | Guide-page top-3 ranking count plus verified brand citations in AI engines. Outcome-anchored, both measurable monthly. |
| Net Promoter Score (composite) | Same anti-pattern as the utility above. Our score was rising while our top-decile customers grew quieter. | Segmented NPS by revenue tier and tenure cohort. The score we report internally is now four numbers, not one. |
Over the years we cut scores of KPIs off our own scorecard, and our scorecard today carries roughly a dozen measures, down from more than thirty a few years ago. Net revenue retention is up over the same window. Two events being correlated is not proof of causation, and I will not pretend it is. But the subtraction discipline has not hurt us, and the number of measures we carry is now far closer to what the platform data says works than to what the Balanced Scorecard rollout I helped sell years ago recommended.
Customers who got the lean discipline right
This is not theory for our customers either. Cities like Durham, North Carolina run public, owner-driven scorecards on ClearPoint. The operating system is simple: a short list, real owners, and a public dashboard that cannot be quietly retired when a number stalls. The public visibility is the accountability mechanism. The discipline is measuring less, more rigorously, with a cadence someone owns.
The pattern holds at the state level too. The Washington Department of Licensing, a state agency serving roughly six million residents, used ClearPoint to pull more than 150 tracked measures down to the critical few its leadership actually reviewed. Fewer measures, real ownership, a fixed review cadence. That is the entire move.
The subtraction discipline
The data is clear enough that I will state the recommendation plainly, before the next OKR rollout or Balanced Scorecard refresh on your calendar:
- Five to nine strategic goals. Not the Balanced Scorecard default of sixteen to twenty.
- Roughly nine to eleven measures. Not three-to-five key results per objective, multiplied across every team.
- One real owner per measure — updated within 30 days, with context, not just a number. This single field roughly doubles your odds of finishing what you start.
- Quarterly review for measures, monthly for projects. And watch the December cliff: annual deadlines bunch hard at year-end, which is where late-cycle execution goes to die.
- Pair every lagging KPI with a leading one: pipeline ahead of revenue, milestone hit rate ahead of completion, engagement score ahead of turnover.
- Pair every gameable KPI with a counter-metric. Time-to-Fill with quality of hire. Training hours with test-pass rate. Closed tickets with substantive-reply time.
- Build a subtraction step into your annual planning cadence. If you do not have one, the framework expands by default. Mine did, for ten years, before I built one.
The plans that complete what they start do not have a secret KPI. They have a smaller list, real owners, and a strategy office willing to enforce the 30-day update cadence. That is the entire story behind the gap between the organizations that finish their strategy and the ones that do not.
Download the ClearPoint Strategic Planning Report
See how ClearPoint tracks your KPIs — book a demo
Common questions
What is the KPI paradox?
The KPI paradox is the documented pattern that organizations are tracking more KPIs than ever — the median strategic plan on the ClearPoint platform carries well over a hundred measures — while completing less of their strategy. In our data, initiatives with a named owner complete at 2.5× the rate (21.5% versus 8.7%), and ClearPoint shows which strategic objectives still need an active owner. The cause is not technology, talent, or budget. It is that almost no measurement framework ships a subtraction discipline.
How many KPIs should an organization actually track?
Far fewer than most frameworks recommend. The Balanced Scorecard default of four perspectives times five-to-seven objectives times three-to-five measures lands at 60 to 105 measures; the plans that actually complete what they start carry a fraction of that. A practical target is five to nine strategic goals and roughly nine to eleven measures, each with a single real owner. That is roughly a quarter of what a default Balanced Scorecard or OKR rollout produces.
Why do Balanced Scorecard and OKR defaults break at scale?
Both frameworks add structure without adding a counterweight to remove anything. Four perspectives times several objectives times several measures compounds quickly; a typical OKR rollout of five objectives times three-to-five key results, multiplied across every team, lands at well over a hundred active key results. The frameworks expand over time because there is no scheduled subtraction step, and execution suffers as the element count climbs.
What does it mean that a measure has "no active owner"?
It means no specific person is responsible for keeping that measure current and acting on it. In the ClearPoint data, ownership is the single most predictive field we track — and ClearPoint shows which measures and objectives still need an active owner. Objectives with an active owner are rated on-track about 23.6% of the time versus about 9.7% without one, a difference from one field. Assigning a real owner who updates the measure on a 30-day cadence is the highest-leverage change most teams can make.
How are operational KPIs different from strategic KPIs?
Operational KPIs measure how the work runs day to day (response times, ticket volume, cycle time, output). Strategic KPIs measure whether the organization is advancing its long-term position (new business model adoption, transformation completion, customer-cohort retention). Operational measures tend to survive because they piggyback on data sources that already exist; strategic measures that require manual surveys or stakeholder coordination are far more likely to be abandoned. The practical recommendation: prefer measures whose data source already lives in a system you operate.
How often should KPIs be reviewed?
Measures monthly, with a quarterly executive recalibration. Projects monthly. Milestones tied to the natural rhythm of the work, not the fiscal calendar. Watch for deadline bunching at year-end — annual initiative deadlines cluster heavily in December, and that bunching is a major driver of late-cycle execution failure.
What is the difference between leading and lagging KPIs?
Lagging KPIs measure what already happened — revenue, churn, completion rate — the rearview mirror. Leading KPIs predict what is coming — pipeline, milestone hit rate, engagement score — the windshield. The most effective scorecards pair every lagging KPI with at least one leading indicator that signals one to three months earlier. Scorecards that carry only lagging measures lose the ability to course-correct in time.
Doesn't every KPI eventually get gamed (Goodhart's Law)?
Often, which is why every gameable KPI on a serious scorecard ships with a counter-metric. Time-to-Fill ships with quality of hire. Training hours ship with test-pass rate. Tickets closed ship with time-to-first-substantive-reply. The fix is not removing the KPI — it is adding the counter-metric to the same conversation so the trade-off becomes visible.
What did ClearPoint Strategy kill from its own scorecard?
Monthly Active Users (decoupled from net revenue retention), Tickets Closed per Week (gamed in real time), Number of Blog Posts Published (an activity substitute with no outcome signal), and the composite Net Promoter Score (which hid trouble among our best customers). We replaced them with Active Owners per Plan, Time-to-First-Substantive-Reply, guide-page ranking count, and segmented NPS. Our scorecard went from more than thirty measures to roughly a dozen, and net revenue retention rose over the same window.
What is the single most important change to make to a KPI program?
Build a subtraction step into the annual planning cadence — the framework expands by default if you do not. After years of platform data and more than a thousand implementations, the highest-leverage change is not adding a metric; it is deciding which metric no longer deserves attention and removing it. A lean list is the destination; the subtraction discipline is the mechanism that keeps you there. And give every measure a real owner — that one field roughly doubles the odds it ends up on-track.








%2520Strategy%2520vs%2520Tactics%2520Differences%252C%2520Examples%2520%2526%2520How%2520to%2520Track%2520Both.jpg)



















.png)




.png)


















.webp)















%2520OKRs%2520vs%2520KPIs%2520Breaking%2520Down%2520the%2520Difference.jpg)