By Ritvars Mētra, Founder of ReadinessCompassThere is a particular kind of dashboard that looks impressive and does almost nothing.
You’ve seen it. Coloured tiles. Trend arrows. A readiness score. Maybe a heat map by region. A few survey questions turned into percentages. “Awareness: 74%.” “Manager confidence: amber.” “Training completion: 91%.” Everyone nods. Someone says the results are “interesting.” Another person asks whether the chart can be filtered by function. The meeting moves on.
No decision is made.
This is the quiet failure of many adoption surveys. They produce information, but not action. They measure sentiment, but not consequence. They show the organisation something about itself, then politely avoid asking what leaders must now do differently.
Why adoption surveys often fail to create action
That’s how readiness metrics become decorative.
Not useless exactly. Decorative metrics can still create the feeling of control. They reassure programme teams that the “people side” is being monitored. They give sponsors something to look at. They help change managers prove that work is happening. But when the launch comes closer, when users are confused, when managers are inconsistent, when the old workaround remains faster, the dashboard’s beauty becomes faintly embarrassing.
The problem, as Harvard Business Review has long observed, is that organisations tend to measure what is easy rather than what drives results. Training attendance, communication reach, and sentiment scores are countable. Adoption behaviour—whether someone can actually perform the new task under pressure—is harder to quantify, so it often gets replaced by softer indicators.
A readiness metric should not be a wall ornament. It should be a trigger. It should trigger a decision, an intervention, an escalation, a delay, a sponsor action, a training redesign, a capacity discussion, or—sometimes—the uncomfortable admission that the organisation is not ready under the current assumptions.
What readiness metrics should help leaders decide
Most teams begin with the data they want to collect. “Let’s ask about awareness, confidence, leadership support, training, resistance, and communication.” Sensible enough. These are familiar categories. They also lead to familiar dashboards. The problem is not the categories; the problem is that the survey is designed as a measurement exercise rather than a decision mechanism.
A better starting point is: what decisions must this readiness data help leaders make?
Should we proceed with go-live? Which business area needs additional support? Where do leaders need to intervene personally? Which managers are not ready to lead their teams through the change? Where is the planned adoption target unrealistic? Which user groups need more practice before launch? Where is the process design creating avoidable resistance? Which old behaviours are likely to survive unless governance changes?
Those are useful questions. They have teeth. Good readiness metrics should help leaders:
- see where adoption risk is concentrated,
- understand whether managers can reinforce the change,
- identify capability and confidence gaps,
- decide where intervention is needed,
- track whether readiness is improving before go-live.
The difference between decorative dashboards and decision metrics
Once the decision is clear, the metric can be designed backwards. If leaders need to decide whether go-live is safe, the survey cannot only ask whether employees “feel informed.” It needs to test clarity, ability, confidence under real conditions, manager support, capacity, and unresolved blockers. If leaders need to decide where to allocate floor-walker support, the data must be segmented by user group, location, process area, and risk level. If leaders need to decide whether managers are ready, don’t ask employees vaguely whether leadership is supportive; ask whether their direct manager can explain what changes, what stays, what is expected, and where to get help.
Specific decision. Specific metric. This sounds obvious, but organisations violate it constantly.
They measure what is easy. Training attendance. Email open rates. Portal visits. Average survey score. Number of champions. Number of briefings held. These are not wrong. They are just incomplete, and sometimes dangerously soothing. A team can complete all change activities and still not change the work.
Training completion is especially seductive. It is clean. It is countable. It travels well in steering decks. Yet training completion tells us who attended, not who can perform. In a software implementation, that difference can be expensive. In an AI implementation, it can be risky. Someone may have completed AI training and still overtrust outputs, avoid the tool entirely, ignore review standards, or use it only for low-value tasks while the real process remains untouched.
So ask the harsher question: what would prove readiness in behaviour? Not in mood. Not in attendance. Behaviour.
Can users complete the new workflow without help? Can managers explain the new rules without contradicting each other? Are old spreadsheets being retired or merely hidden? Are support tickets showing confusion in one process step? Are teams using the new CRM data in actual pipeline reviews? Are employees applying the AI review standard before outputs enter customer-facing work? Are approval exceptions decreasing? Are data fields completed correctly, not just completed?
A readiness survey can’t answer all of this by itself. That’s another trap. Surveys are useful because they capture perception, confidence, clarity, and local concerns at scale. But readiness is not only what people say. It is also what they do, what managers reinforce, what systems reveal, and what workarounds survive.
A serious readiness metric system triangulates. Survey data. System usage. Process audits. Training practice results. Manager check-ins. Support-ticket themes. Workshop observations. Adoption KPIs. Open comments. Maybe even informal signals from super users who know where the real trouble is hiding. The survey is one instrument in the orchestra. It is not the whole orchestra.
How to design adoption survey questions that expose risk
Still, adoption surveys can be powerful if they are designed with decisions in mind. The wording matters. Many readiness surveys ask questions that create pleasant ambiguity: “I understand the change.” “I feel prepared.” “Leadership communicates effectively.” Fine, but prepared for what? Which change? Which task? Which leadership? Under what working conditions?
A better question is anchored in the actual adoption behaviour:
“I know which reports will be retired and which report I must use after go-live.”
“My manager has explained how the new process affects our team’s priorities.”
“I can complete the new customer handover process without using the old template.”
“I know what to do if the AI-generated response looks plausible but unsupported.”
“I have enough time before go-live to practise the tasks I will perform weekly.”
“I believe leaders will stop accepting the old workaround after launch.”
That last one is deliciously uncomfortable. It tests reinforcement, trust, and leadership consistency in one line. It also reveals a common adoption risk: employees may understand the new way but not believe the organisation will enforce it. They have seen too much corporate theatre. They wait.
Metrics should also avoid the tyranny of averages. Average readiness is a dangerous friend. It makes weak spots disappear. A programme can show an overall readiness score of 78% while one critical function sits at 42%, one region has not completed role mapping, and one stakeholder group quietly controls the data quality needed for everyone else’s success. Adoption does not fail on average. It fails in specific places.
Segment the data. By stakeholder group, role, process, country, manager population, business unit, adoption-critical behaviour. Not endlessly—there is such a thing as slicing the data until it becomes confetti—but enough to see where decisions are needed. A readiness dashboard should help leaders say, “This is where we act next,” not merely, “This is where the colours are different.”
How to turn survey results into leadership actions
Thresholds help, but only if they are tied to action. For example, if fewer than 70% of impacted users can identify the new process owner, then process ownership must be clarified before training continues. If manager confidence drops below a defined level in a high-impact group, the sponsor must run a manager briefing within five working days. If survey comments show repeated concern about workload, the programme must review capacity or adjust sequencing. If usage data after go-live shows low adoption in a region that scored high on readiness, the issue must be investigated rather than celebrated as a “survey anomaly.”
The threshold is not the point. The response is.
This is where leadership action enters. A metric without an owner is a mood ring. It changes colour; nobody has to do anything. Every readiness metric that appears in a leadership dashboard should have three things attached: an accountable owner, a decision rule, and a review rhythm.
Converting adoption survey results into action typically follows five steps:
- Segment the results by audience, workstream, or business unit.
- Identify the weakest readiness signals — where capability, confidence, clarity, or reinforcement is lowest.
- Translate each signal into a business risk using plain language: “If this is not addressed before go-live, the likely consequence is…”
- Assign an owner and intervention — who will act, what they will do, and by when.
- Recheck after the intervention — did the metric move? If not, escalate or redesign the intervention.
Without that discipline, the dashboard becomes a reporting ritual. People update it because governance requires it. Leaders look at it because it is on the agenda. The numbers move, the programme continues, and the same adoption risks appear after launch wearing different clothes.
There is also a timing problem. Readiness metrics are often collected too late, close to go-live, when most meaningful decisions are politically expensive. By then, the launch date is public, the training plan is booked, the sponsor has promised benefits, and any serious readiness finding is treated as inconvenient noise.
A readiness survey two weeks before go-live is not useless. It can help target hypercare. But it is too late to shape adoption properly. The better pattern is baseline, pulse, action, pulse again—baseline early enough to influence design and planning, pulse after key communication or manager briefings, pulse after training practice, pulse before go-live, and pulse after go-live when behaviour is visible. Not too many surveys—we don’t need to harass people into readiness—but enough to see whether interventions are working.
Readiness signal
Business risk
Leadership action
Recheck
Example: readiness metrics become useful when each weak survey signal is translated into an owned leadership action and checked for impact.
What to show in a readiness dashboard
Trend matters more than snapshot. A single amber score tells us little. A declining confidence score among managers after training tells us something sharper: perhaps the training revealed unresolved complexity. A rising awareness score with flat ability score suggests people know what is coming but cannot yet perform it. High willingness with low capacity means the organisation is not resisting; it is overloaded. Low trust with high understanding means the message landed but leadership credibility did not.
This is where metrics become diagnostic.
And diagnosis should lead to different medicine. If awareness is low, communicate. If ability is low, practise. If capacity is low, reprioritise. If trust is low, leaders must show up differently. If process fit is poor, redesign. If reinforcement is weak, change governance and manager routines. If old behaviours remain rewarded, stop pretending the problem is training.
One of the worst habits in change management is prescribing communication for every readiness gap. Communication is important, but it cannot solve role ambiguity, lack of time, broken process design, weak sponsorship, poor data quality, or contradictory incentives. A decision-oriented dashboard makes this visible by linking each metric to a plausible action category.
Think of the dashboard less as a display and more as a decision table:
| Metric | Possible action |
|---|---|
| Users understand why the change is happening | Sponsor message, local Q&A, customer/business case story |
| Users can perform the new task | Scenario practice, job aid, floor support, simplify workflow |
| Managers reinforce the new behaviour | Manager briefing, escalation to functional leader, governance rule |
| Old workaround remains attractive | Retire old tool, change reporting requirement, block parallel process |
| Capacity is insufficient | Reduce concurrent demand, extend transition, add temporary support |
That last category is the one leaders often dislike most because it requires trade-offs. But readiness metrics that never force trade-offs are probably too polite.
The most valuable adoption survey is not the one that tells leaders employees are “mostly positive.” It is the one that forces a leadership team to decide whether they are willing to create the conditions for adoption. Are they willing to protect time for practice? Stop accepting old-process outputs? Challenge influential managers? Add support where the data shows risk? Delay a release if readiness is genuinely too low? Remove a metric that rewards the wrong behaviour?
If the answer is no, the dashboard was never the issue.
Common mistakes when reporting readiness metrics
A nice refinement is to include a “decision requested” field directly on the readiness dashboard. Not hidden in speaker notes. Directly visible:
“Decision requested: confirm whether old customer handover template will be disabled at go-live.”
“Decision requested: approve additional support for warehouse teams during first two weeks.”
“Decision requested: sponsor to address low manager confidence in DACH and Benelux before next pulse.”
“Decision requested: agree whether AI use in draft customer communication is optional, recommended, or mandatory for defined cases.”
This changes the tone. The dashboard stops asking leaders to admire the data. It asks them to govern the change.
Naturally, there are risks. A poorly designed readiness metric can create surveillance anxiety. Employees may fear that honest answers will be used against their managers or teams. Survey fatigue can reduce response quality. Dashboards can oversimplify messy human realities. Leaders may weaponise red scores. Change teams may overfit interventions to noisy data. None of this should be ignored.
Trust matters. Explain why the data is collected. Share what changed because of feedback. Protect anonymity where promised. Avoid ranking teams in a way that humiliates rather than helps. Combine quantitative patterns with qualitative sense-making. Be careful with small samples. Don’t pretend a decimal point makes judgement scientific.
Good readiness metrics are humble. They don’t claim omniscience. They say: here is what we are seeing, here is what it probably means, here is the decision needed, here is how we will know whether the action worked. That is enough.
A simple operating rhythm for adoption decisions
The final test is brutally simple. After the readiness dashboard is reviewed, can you name the decisions made because of it?
If not, it is decorative. If yes—if a sponsor intervened, a training module was redesigned, a release was resequenced, a manager layer was equipped, an old workaround was removed, support was shifted to the right team, or a go-live assumption was challenged—then the survey has done its job.
This approach aligns with what McKinsey calls the “people power of transformations” — the recognition that organisational change succeeds when metrics connect directly to leadership decisions, not just project reporting. Similarly, Gartner has noted that change-ready organisations build adoption measurement into governance, not just into communication plans.
Research from MIT Sloan Management Review reinforces that leadership in transformation is about creating conditions where evidence drives behaviour — not about managing dashboards, but about acting on what the data reveals. And Nielsen Norman Group has documented that poorly designed dashboards fail when they display data without decision context, a pattern familiar to anyone who has watched a readiness presentation that generated nodding but no action.
Adoption surveys should not exist to prove that change management is busy. They should help leaders see where adoption will break, decide what to do, and then check whether the action changed anything.
That is the journey from survey to leadership action. Everything else is dashboard wallpaper.