Strategy Health Metrics: Seven Vital Signs, and What an Abnormal Reading Actually Means

Table of Contents

Key Takeaways

• Health metrics measure the system, not the results: a company can hit its numbers with an unhealthy execution system, and miss them with a healthy one. Conflating the two is the most common error in this category.
• Every abnormal reading needs a differential: falling check-in discipline could be overload, disengagement, or a reporting-cycle artefact. The three call for entirely different responses.
• Confidence and progress are recorded independently in Profit.co: a key result can carry a Completed status while its progress slider sits at 50%. The gap between the two is itself one of the most diagnostic signals available.
• Configuration invalidates naive comparison: confidence percentages behind each status are customizable, and status propagation can be disabled per status, so the same reported number can mean different things in two departments.
• Trend beats snapshot on every vital sign: Profit.co’s status distribution widget shows how OKRs shift across statuses over time, which carries far more signal than any single-period cut.
• Healthy ranges are practitioner heuristics, not vendor thresholds: the metric definitions here are sourced from Profit.co documentation; the suggested bands are editorial guidance to be replaced by your own baseline within two cycles.

1. What Strategy Health Metrics Are, and What They Are Not

Strategy health metrics measure whether the machinery of execution is functioning. They are deliberately separate from business results, and keeping them separate is the entire point.

An organization can post an excellent quarter with a broken execution system, favourable market, one heroic team, a deal that closed for reasons nobody planned. It can also post a poor quarter with a healthy one, having made good decisions against conditions that did not cooperate. If health metrics and outcome metrics are mixed into a single view, both become uninterpretable: the good quarter conceals the broken system until the market turns, and the bad quarter triggers a redesign of machinery that was working correctly.

The distinction in practice

• Outcome metrics answer whether the business is winning. Revenue, retention, margin, market share.
• Health metrics answer whether the system that pursues those outcomes is intact. Are commitments owned, updated honestly, funded coherently, and reviewed in time to act?

Health metrics are leading; outcome metrics are lagging. That is why a strategy function should be measured on the first and the business on the second, a boundary explored further in the context of why most enterprise OKR programs fail in year two, where systems that were never healthy are carried by a strong first year and collapse in the second.

2. The Diagnostic Principle: One Reading, Several Causes

The reason most health metrics fail to change anything is that organizations treat a reading as a conclusion. It is not. It is a symptom, and symptoms are ambiguous.

Check-in discipline drops from 91% to 68% in a department. That single fact is consistent with at least three distinct situations:

• The team is overloaded and updating is the first thing to go. The system is fine; the capacity is not.
• The team has disengaged because their objectives feel disconnected from their real work. The capacity is fine; the objectives are wrong.
• A reporting-period boundary, a holiday, or a reorganisation moved the denominator. Nothing is wrong at all.

Each calls for a different response, resourcing, re-authoring, or nothing. Acting on the wrong one costs credibility, and acting on the third costs it fastest, because the team knows nothing was wrong.

The three-step discipline

• Read the trend, not the point. One period is noise. Three consecutive periods moving in the same direction is a signal.
• Check the configuration before the conclusion. Section 11 covers the specific settings that can produce an abnormal reading with nothing abnormal behind it.
• Corroborate with a second vital sign. No single metric is diagnostic on its own. Section 12 works through three cases where two readings together identify a cause that neither identifies alone.

3. Vital Sign 1, Check-in Discipline

Measures: participation integrity, whether the data underneath every other metric is being maintained.

Profit.co breaks check-in behaviour into Missed, Late, On time, and Pending for the current week, with Score, Count, and Percent available for previous weeks and a date-range filter across the quarter. Data is viewable by Individual, Department, and Team, with an owners tab naming exactly which key results are outstanding. A separate Cockpit widget highlights the trend of users completing check-ins on time.

Suggested healthy band

On time above 85%, Missed below 5%, sustained across three consecutive weeks. Treat these as starting heuristics rather than standards, replace them with your own baseline once you have two cycles of history.

Differential

• Overload. Late rises while Missed stays low. People are still updating, just behind. Corroborate with capacity data.
• Disengagement. Missed rises rather than Late. People have stopped rather than slipped. Corroborate with alignment coverage, disengagement usually tracks objectives that never connected to real work.
• Artefact. The whole organization moves at once. A holiday, a cycle boundary, or a change in who is counted. Check the denominator before escalating.

This is the first vital sign because it is the one all the others depend on. If check-in discipline is poor, every downstream metric is being computed from partial data, and the correct action is to fix the input before interpreting the outputs.

4. Vital Sign 2, Confidence-to-Progress Divergence

Measures: forecast honesty, whether owners’ stated confidence matches their measured progress.

This is the least-used vital sign and one of the most informative, because Profit.co records the two quantities independently. Progress is the Value field, a quantitative measure of movement toward the target. Confidence is the status dropdown, an owner’s judgement about whether the target will be met in time. Profit.co’s documentation is explicit that progress and confidence are treated independently, a key result can carry a Completed status while its progress slider reads 50%.

Each status maps to a configurable confidence percentage. In the default configuration On Track corresponds to a high confidence value, At Risk to a middling one, and In Trouble to a low one, and an objective’s status is calculated from the average confidence percentage of its key results, an average of 40%, for instance, falls in the At Risk band.

What divergence means

• High confidence, low progress. The owner expects a late surge. Legitimate for back-loaded work; a warning sign when the progression model is Linear and no surge is planned.
• Low confidence, high progress. The owner knows something the metric does not, a dependency about to fail, a number that will reverse. This is the most valuable reading on the entire panel and the one most often overridden by managers who look only at progress.
• Confidence that never moves. An owner reporting On Track every week regardless of progress is not forecasting. They are defaulting. Flat confidence across an entire quarter is a data-quality problem disguised as good news.

Profit.co also supports an optional end-of-cycle Projection Status, enabled from Settings under Status Tracking, which adds a field at check-in for forecasting the expected end-of-cycle outcome alongside the actual status. Where it is enabled, projection-versus-actual divergence is a sharper version of this same signal. Check-in mechanics are documented in the key result check-in guide.

5. Vital Sign 3, Say-Do Ratio

Measures: commitment reliability, whether what a team promises predicts what it delivers.

Where check-in discipline is a weekly leading signal, Say-Do is a cycle-over-cycle trailing one. Say-Do Ratio configuration allows a cutoff for KPIs and initiatives and a separate cutoff for company and department OKRs, 80% is a common starting point, with check-in status resolving to Not Started, On Track, In Trouble, or Exceeded. Calculation runs at parent level only or at parent plus sub level.

Differential

• Chronic over-commitment. Say-Do consistently below band while progress is respectable. The team is ambitious and the targets are unrealistic. This is a target-setting conversation, not a performance one.
• Sandbagging. Say-Do consistently at or above 100%. Targets are being set below capability. Corroborate with stretch performance, a team that always hits exactly its number is calibrating, not stretching.
• Genuine delivery difficulty. Say-Do falling while check-in discipline also falls and confidence drops. Three signals moving together is the one case where the reading is unambiguous.

Say-Do is the vital sign most vulnerable to being read as a performance rating. It is not one. A team with a low ratio and honest forecasting is healthier than a team with a perfect ratio and targets set to guarantee it.

6. Vital Sign 4, Alignment Coverage

Measures: structural integrity, whether the cascade actually connects.

Two numbers: the share of corporate objectives with at least one aligned objective beneath them, and the share of departmental objectives with a valid parent. Profit.co’s Cockpit surfaces this in its Plan phase, which evaluates Ownership, Accountability, and Alignments , with an Organization Alignment to Corporate Objectives view showing assigned and aligned key results, their progress and status, and planned versus actual progress with confidence percentage.

Suggested healthy band

Corporate coverage at 100%, anything less is a named strategic priority nobody is working on. Orphaned departmental objectives below 10%, with the caveat that a small number is healthy: teams doing necessary work the strategy did not anticipate is information, not failure.

Differential

• Architecture failure. Coverage gaps cluster at one level or in one function. The cascade design did not reach there.
• Timing artefact. Coverage is low early in a cycle and resolves within two weeks as departments complete planning. Measure at cycle lock, not cycle open.
• Strategic ambiguity. Coverage is low against one specific corporate objective across every department. Usually the objective is not actionable, and the fix is upstream in authoring rather than in the cascade.

That third case is worth dwelling on, because it is regularly misdiagnosed as a discipline problem. If no department has aligned to a corporate objective, the most likely explanation is that nobody knows what they would do about it. Guidance on the underlying architecture is in cascading OKRs across the enterprise without losing strategic alignment.

7. The Seven Vital Signs Reference Table

A reference panel. The bands in column three are practitioner starting points, not vendor-published thresholds, replace them with your own baseline as soon as you have two cycles of history.

Read the final column as the working content of the table. Each of these misdiagnoses is common enough to be the default reading, and each one sends the response in a direction that makes the underlying problem worse rather than better. The instrumentation that produces these readings sits across Profit.co’s insights, analytics, and reports, including the Alignments View, Status Map View, Bowler Chart, Check-in Discipline, and Say-Do Ratio.

8. Vital Sign 5, Status Distribution Shift

Measures: portfolio momentum, the direction the whole objective set is moving, rather than where any one objective sits.

Profit.co’s Cockpit includes a widget that displays how OKRs are distributed across On Track, At Risk, In Trouble, Completed, Not Started, and Archived and, critically, how those shift over time. The temporal element is what makes this a vital sign rather than a status report.

Why the snapshot misleads

A distribution of 60% On Track, 25% At Risk, 15% In Trouble looks identical in two organizations with opposite health. In the first, the same objectives have held those positions all quarter, stable, predictable, manageable. In the second, objectives are churning between bands weekly, and the stable-looking aggregate is hiding an execution environment nobody can plan against. Only the shift view distinguishes them.

Differential

• Healthy convergence. Net movement toward On Track and Completed as the cycle progresses, with churn declining. This is what a working system looks like.
• Late-cycle optimism. A sudden shift toward On Track in the final weeks without corresponding progress movement. Corroborate against Vital Sign 2, this is usually confidence being adjusted to match the desired reporting outcome rather than the underlying reality.
• High churn, flat aggregate. Objectives moving between bands repeatedly while the totals stay constant. Indicates unstable targets or volatile dependencies, and it is invisible in any snapshot.

9. Vital Sign 6, Portfolio-to-Objective Linkage

Measures: funding integrity, whether money is going where the strategy says it should.

One number: funded initiatives with no link to an active objective. In most portfolios it is non-zero, and in most organizations nobody has looked at it this year. It is the cheapest strategic finding available and requires no new instrumentation to produce.

Profit.co’s strategic portfolio management module links projects directly to company OKRs with fund-to-goal traceability, so the unlinked count is a query rather than an audit. The design principle it enforces is that no dollar is approved without an objective behind it.

Differential

• Genuine drift. Initiatives approved against priorities that have since changed, still running on inertia. The action is cancel, re-scope, or re-link, and re-link is the right answer more often than people expect.
• Linkage hygiene. The work does serve an objective; nobody recorded the connection. Real problem, but a data problem rather than a strategic one, and far cheaper to fix.
• Legitimate run-the-business work. Compliance, maintenance, infrastructure. This should be classified as such rather than forced into a strategic objective, which is how objective sets become diluted with operational necessities.

The third case matters. A portfolio where every initiative maps to a strategic objective is usually one where strategic objectives have been written to cover work that was happening anyway. That is a different pathology with the same clean-looking number, and it is examined in portfolio optimization as the execution gap most companies miss.

10. Vital Sign 7, Benefit Realization Rate

Measures: outcome integrity, whether completed work delivered the value it promised.

The longest feedback loop on the panel and the one most commonly missing entirely, because delivery systems stop measuring at handover. Profit.co’s Value Realization Office compares actual benefits against plan, translates earned value into ROI, and continues tracking value past project close, with IRR, NPV, and Payback Period recalculating automatically as projects progress.

Differential

• Execution shortfall. The initiative delivered less than scoped. Delivery data will corroborate.
• Business case inflation. The initiative delivered fully and the promised benefit never existed. Delivery data will look clean. This is the most important finding this vital sign produces and the one most likely to be attributed to the delivery team instead of the approval process.
• Measurement lag. The benefit is real and has not landed yet. Distinguishable only if the business case specified a realization timeline, which is itself a useful thing to discover.

Read this immediately before a funding decision rather than at quarter end out of habit. Its entire purpose is to change the next allocation, and a benefit realization report read three weeks after the budget closes has no mechanism by which to matter.

11. Configuration Caveats That Invalidate a Reading

Before escalating any abnormal reading, check whether configuration explains it. Four settings in particular can produce an alarming number with nothing wrong behind it, or, worse, a reassuring number with something wrong.

Confidence percentages are customizable

The confidence value behind each status is editable under Settings, OKRs, Authoring, Statuses, and organizations can add entirely new statuses with their own confidence levels. Two departments reporting the same average confidence may be using different bands. Comparison across units is only valid once the configuration is confirmed identical.

Status propagation can be disabled per status

Profit.co’s objective status propagation calculates an objective’s status from the confidence percentages of its key results, but propagation can be disabled for specific statuses, and disabled statuses are excluded from the calculation entirely. Disabling On Track, for example, causes an objective to be scored only on its struggling key results and appear far worse than it is. This is the single most common cause of an inexplicable objective status.

Progress and confidence move independently

Because the two are recorded separately, a portfolio can show healthy progress and deteriorating confidence at the same time without contradiction. Any health panel that surfaces only one of them is blind to half the signal.

Weighting changes what an average means

Where key results are weighted, objective progress is a weighted contribution rather than a simple average. Profit.co’s own guidance is to check whether an objective’s key results are weighted before interpreting its position in any aggregate view. An objective can appear to be failing because its single heaviest key result is behind, while four lighter ones are complete.

12. Three Differential Diagnoses Worked Through

Case 1: Discipline falling, progress steady

Case 2: Everything green, confidence flat

Case 3: Portfolio healthy, benefits missing

In all three cases one metric alone produced the wrong read and two together produced the right one. That is the argument for a panel rather than a headline number, and it mirrors what separates functioning from stalling programmes in the OKR programme health check.

13. Gaming, and Five Numbers That Look Like Health But Are Not

Any metric attached to consequences will be optimized. Health metrics are unusually vulnerable because they are process measures, which are cheaper to satisfy directly than the outcomes they proxy.

The three gaming patterns to expect

• Check-in theatre. Updates submitted on time with no substantive change. Detect by pairing discipline with progress velocity, perfect discipline and static progress is the signature.
• Status defaulting. On Track selected every week regardless of reality. Detect through confidence variance: a distribution with no spread is not a forecast.
• Target deflation. Say-Do improves because commitments shrink. Detect by tracking target ambition across cycles alongside the ratio.

The structural defence is to never attach individual consequences to a process metric. Check-in discipline used to evaluate people stops measuring participation and starts measuring compliance, at which point it has lost the diagnostic value that made it worth collecting.

Five numbers that look like health but are not

• Number of objectives created. Measures activity. Often inversely related to focus.
• Overall completion percentage. Averages across initiatives of different shapes and stakes into a figure that moves slowly and conceals everything interesting.
• Login frequency. Measures tool usage, not execution health. A team can log in daily and decide nothing.
• Number of check-ins submitted. Volume without quality. The discipline metric is about timeliness and coverage, not count.
• Percentage of objectives marked On Track. Only meaningful alongside its trend and its confidence distribution. On its own it is the easiest number on the panel to manufacture.

14. Building the Panel: Baseline, Band, Cadence

Step 1, Baseline before you set any target

• Measure all seven for one full cycle without publishing thresholds or attaching any consequence.
• Record the natural variance. A metric that swings 15 points week to week in your organization cannot carry a 5-point threshold.
• Discard the suggested bands in Section 7 as soon as your own data disagrees with them. They are starting points, not standards.

Step 2, Set bands from your own distribution

• Set the action threshold outside normal variance, so crossing it means something.
• Define what each band triggers before anything crosses it. A threshold with no defined response is a notification, not a control.
• Where the platform can enforce it, automate the trigger rather than relying on someone noticing, threshold alerts on projects, portfolios, tasks, and milestones can fire through the Action Center, email, a dashboard badge, or a webhook.

Step 3, Assign cadence and owner

• Weekly: check-in discipline, confidence divergence. Monthly: status distribution, portfolio linkage. Per cycle: Say-Do, alignment coverage. Quarterly and pre-funding: benefit realization.
• Name one owner for the panel. Metrics owned by everyone are reviewed by nobody.
• Anchor the readings to an existing review rather than creating a new meeting, using the meetings module so decisions attach to the goals they affect.

Step 4, Protect the panel from drift

• Never attach individual performance consequences to a process metric.
• Re-baseline annually. Bands set against a 2,000-person organization do not survive its growth to 4,000.
• Audit for vital signs that have sat inside band for four consecutive quarters, they are either measuring something invariant or measuring it too coarsely.

Profit.co reports most customers complete setup and run their first cycle within two to four weeks, so a baseline cycle followed by a banded cycle puts a working health panel roughly one quarter out. Context on the wider platform decision sits in choosing the right strategy execution platform, and role-level framing on Profit.co’s hub for strategy and transformation leaders.