Dashboard Delirium: When More Data Makes Leaders Strategically Deaf

tl;dr: Leaders do not become more strategic simply because they can see more numbers. Dashboards help when they create shared orientation around a real strategic question, agreed signals, and course correction. They create dashboard delirium when proxy metrics, AI-generated reports, and fragmented information streams become emotional insurance against judgment. The antidote is not to reject data. It is to decide what matters, assign ownership, pair every metric with a listening channel, and ask what the number cannot hear.There is a particular kind of dashboard that makes leaders feel wonderfully responsible.Everything has a number. Every number has a trend line. Every trend line has a color. The weekly review opens, the screens glow, and for a brief moment the organization looks knowable.Then someone asks what should change.Silence.Not because the team lacks data. The data is everywhere: in the BI tool, the CRM, the product analytics suite, the finance model, the customer-success dashboard, the AI-generated weekly summary, the board deck, the team scorecard, and that spreadsheet nobody admits is still the real operating system.The room is not under-informed.It is strategically deaf.That phrase is a metaphor, not a diagnosis. Nobody has lost the physical ability to hear. What disappears is the leadership capacity to hear context, exceptions, weak signals, customer reality, frontline judgment, and tradeoffs that do not fit neatly inside the chart.This is dashboard delirium: the feeling of control created by seeing more and more signals while losing the ability to decide which ones deserve authority.### Dashboards are not the villainPrecision matters here, because lazy arguments are always waiting for a chair in the meeting.Dashboards are useful.<a href="https://cisr.mit.edu/publication/20220101DashboardingWeillWoerner" target="blank" rel="noopener noreferrer">MIT CISR’s 2022 research on dashboarding</a>, based on its 2019 Top Management Teams and Transformation Survey of 1,311 companies, found that organizations with top-quartile dashboard effectiveness outperformed bottom-quartile companies across internal and external performance measures. The important part is not that charts have magical properties. MIT CISR framed effective dashboarding as a way for people to compare progress against agreed metrics and make course corrections together.That is the healthy version.A good dashboard gives an organization shared orientation. It makes assumptions visible. It connects current signals to a strategy people have actually agreed to pursue. It helps leaders notice drift early enough to respond.A bad dashboard does something subtler.It creates the emotional experience of knowing.That is where the danger begins.### When every number gets a microphoneThe leader is not drowning because the dashboard is wrong.The leader is drowning because every metric has been allowed to speak at the same volume.Revenue speaks. Churn speaks. Pipeline speaks. Engagement speaks. Ticket volume speaks. Cycle time speaks. Utilization speaks. Forecast accuracy speaks. Employee sentiment speaks. NPS speaks. AI adoption speaks. AI output volume speaks.The system does not whisper, "By the way, only three of us matter for the decision you are making today."It just keeps speaking.AI can help it speak faster. That is useful when the organization has a decision architecture. It is exhausting when the organization only has reporting capacity.Information overload is not a poetic complaint from people who dislike dashboards. A 2023 <a href="https://www.frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2023.1122200/full" target="blank" rel="noopener noreferrer">Frontiers in Psychology review</a> describes information overload as a problem studied across disciplines, commonly tied to information exceeding processing capacity and associated in prior empirical work with strain, burnout, productivity loss, performance loss, and decision-quality problems.That does not prove your executive dashboard is harming your strategy. It does support the more modest point: more information stops helping when it outruns interpretation.The modern workday already has a head start on that problem. Microsoft’s 2025 <a href="https://www.microsoft.com/en-us/worklab/work-trend-index/breaking-down-infinite-workday" target="blank" rel="noopener noreferrer">Work Trend Index special report</a> reported, using Microsoft 365 telemetry and survey data, that employees using Microsoft 365 were interrupted on average every two minutes during core work hours by meetings, emails, or notifications, and that 48 percent of employees and 52 percent of leaders said work felt chaotic and fragmented.Treat that as vendor-shaped directional evidence, not a census of all work.Still, it names something most leaders already feel. Before we add another AI-generated executive summary, another daily metric digest, and another agent that "proactively surfaces insights," the room is already noisy.Somewhere, a dashboard is now producing a weekly recap of another dashboard.Naturally, it has a tasteful icon.### The proxy becomes the thingThe oldest trap in measurement is still the one leaders fall into with the most professional posture.The measure becomes the work.<a href="https://ideas.repec.org/a/eee/epplan/v2y1979i1p67-90.html" target="blank" rel="noopener noreferrer">Campbell’s 1979 paper on social indicators</a> is the classic warning label here: when quantitative indicators become heavily used for decisions, they can become vulnerable to corruption pressure and can distort the processes they were meant to monitor. You do not need to turn that into a universal law to see the relevance.If customer health is measured by login frequency, teams may optimize login frequency.If productivity is measured by completed tickets, teams may complete tickets.If AI adoption is measured by number of prompts, people may prompt more.Congratulations. The number moved.The question is whether the business did.Management-accounting researchers use an even sharper word for this: surrogation. In <a href="https://scholarsarchive.byu.edu/facpub/8217/" target="blank" rel="noopener noreferrer">"Surrogation Fundamentals: Measurement and Cognition"</a>, Paul Black, Thomas Meservy, William B. Tayler, and Jeffrey O. Williams describe the tendency for managers to lose sight of strategic constructs and behave as though the measures are the constructs themselves. Their paper also describes surrogation as a nonconscious process.That matters. The problem is not only cynical gaming. It is not always someone sitting in a dark room asking how to make the metric look good while the customer quietly suffers.More often, competent people adapt to what the organization keeps rewarding, reviewing, and coloring red.The dashboard becomes the loudest representation of reality in the room. Then, gradually, it becomes reality’s replacement.### Data-driven is not data-obedientThere are two bad responses to this.One is to worship the dashboard. The other is to sneer at it. Both avoid the harder work.Michael Luca and Amy Edmondson have made this point clearly in Harvard Business Review’s work on <a href="https://hbr.org/2024/09/where-data-driven-decision-making-can-go-wrong" target="blank" rel="noopener noreferrer">where data-driven decisions go wrong</a> and in the related <a href="https://hbr.org/podcast/2025/03/the-right-way-to-make-data-driven-decisions" target="blank" rel="noopener noreferrer">HBR On Strategy episode</a>. The issue is not simply whether leaders use data. It is whether they interpret the data in relation to the decision, its context, the quality of the evidence, what was measured, what was omitted, and what the result can or cannot say.That is the difference between being data-driven and being data-obedient.A data-driven leader asks better questions because the evidence exists.A data-obedient leader treats the evidence as permission to stop thinking.That distinction will matter more as AI increases the volume of analysis leaders can request. McKinsey’s <a href="https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai" target="blank" rel="noopener noreferrer">2026 State of AI survey</a> describes broad AI use across business functions while still focusing on the challenge of translating individual gains into enterprise impact. It also reports that high performers are more likely than others to redesign workflows because of AI use.That does not prove AI is causing dashboard overload. It does suggest a familiar pattern: organizations can expand analytical and reporting capacity faster than they redesign the work, decisions, and accountability around it.Wavestone’s <a href="https://www.wavestone.com/en/news/2024-data-and-ai-leadership-executive-survey-41/" target="blank" rel="noopener noreferrer">2024 Data and AI Leadership Executive Survey</a> points in the same direction from another angle. It reported that 87.9 percent of participants said data and analytics investment was a top organizational priority, while also naming the time and commitment required to integrate data and AI into business processes and culture.Investment can buy more instruments.It cannot, by itself, decide what music the organization is trying to play.### What the number cannot hearEvery dashboard has a hearing range.It can hear what the organization has chosen to instrument. It can hear what teams have defined, tagged, counted, normalized, and allowed into the system. It can hear what happened often enough to become visible in aggregate.It struggles with the rest.It may not hear the customer who renews but no longer trusts you.It may not hear the support agent who keeps fixing the same product ambiguity in private because logging it properly takes longer than helping the person in front of them.It may not hear the sales conversation where the buyer technically fits the ICP but is asking for a future you should not build.It may not hear the frontline workaround that looks inefficient because the official process is pretending reality has better manners.It may not hear the one weird exception that should change your understanding of the system.Qualitative evidence is not automatically superior to quantitative evidence. A vivid anecdote can mislead just as badly as a clean chart.But when leaders remove listening from the operating system, the dashboard becomes a filter that no one remembers choosing.That is how organizations become less strategic while feeling more informed.### The orientation testThe fix is not fewer numbers as an aesthetic preference. The fix is a tighter relationship between strategy, measurement, listening, and decision ownership.Before adding another dashboard, ask four questions.What decision is this meant to improve?If the answer is "visibility," keep pressing. Visibility into what? For whom? By when? Toward which choice? A dashboard with no decision is usually a shrine to anxiety.Which strategic question gives this number meaning?Churn means one thing when the strategy is expansion into enterprise accounts. It may mean another when the strategy is learning from early adopters. AI usage means one thing when the goal is workflow redesign. It means another when the goal is performative modernization with better icons.The metric is not self-interpreting.What does this dashboard fail to hear?Every metric should travel with a listening channel. Pair customer metrics with customer conversations. Pair productivity metrics with employee reality. Pair quality metrics with exception review. Pair AI adoption metrics with examples of changed work, judgment, and consequence.The number should start the conversation, not end it.Who owns the decision when the signal moves?This is where many dashboards confess.If a metric turns red and nobody has authority to change the work, the dashboard is not a management tool. It is an illuminated weather report inside a building with no doors.Someone has to own interpretation, action, and correction.### Fewer signals, sharper judgmentThis connects directly to the previous TBAI argument about safety in accelerated environments. In <a href="https://www.linkedin.com/pulse/safe-choice-dangerous-one-zero-buffer-world-maciejewski-zk8je/" target="blank" rel="noopener noreferrer">The Safe Choice Is the Dangerous One in a Zero-Buffer World</a>, the point was that waiting can feel responsible while quietly increasing learning risk.Dashboard delirium is a cousin of that same false safety.The leader feels safer because every screen is instrumented. Every meeting has a report. Every report has an owner. Every owner has a meeting about the report.But safety is not the same as saturation.In a zero-buffer world, leaders need signals that create contact with reality early enough to act. That requires subtraction as much as collection. It requires the discipline to say, "For this decision, these are the few signals that matter most. These are the signals we will watch as context. These are the signals we will ignore for now, even though they are available and wearing a very serious font."That last part is harder than it sounds.Ignoring available data can feel irresponsible. It can feel exposed. It can feel like admitting you are making a judgment call, which is precisely the thing many dashboards were quietly recruited to avoid.But leadership is not the elimination of judgment through better reporting.Leadership is the visible ownership of judgment under imperfect information.### The leader as noise filterThe next serious leadership capability is not collecting more evidence. Most organizations are already very good at collection.The scarce capability is interpretation.What matters now is the leader who can say:<ul><li>This number is a signal, not the strategy.</li><li>This trend is real, but not decisive.</li><li>This outlier deserves attention before it becomes a pattern.</li><li>This metric improved because the team optimized the proxy.</li><li>This customer story changes how we should read the chart.</li><li>This AI-generated analysis is plausible, but it did not define the tradeoff.</li></ul>That is not anti-data. It is respect for data serious enough not to leave it alone in a room with incentives.The dashboard should help leaders hear. It should make the few vital signals clearer. It should support course correction, shared accountability, and faster learning.When it does that, keep it.When it becomes emotional insurance against deciding, interrogate it.When it crowds out the customer, the frontline exception, the dissenting specialist, the tradeoff, or the question nobody wants to ask, turn the volume down.The dashboard is not the strategy.The strategy is the choice about what matters, what tradeoffs are acceptable, what evidence would change your mind, and who is responsible for acting when reality answers back.More data can make a leader better informed.Only orientation makes that information useful.RobP.S. Which number on your dashboard gets the loudest voice in the room, and what important signal might it be drowning out?If you are a leader or founder trying to move from answer-based authority to responsible orientation, let’s talk. I help organizations design the human judgment, context, and decision structures that keep AI-enabled speed connected to consequence.Find out more at <a href="http://www.robkonrad.com" target="blank" rel="noopener noreferrer">www.robkonrad.com</a> and <a href="https://www.thinkingbeyond.ai" target="blank" rel="noopener noreferrer">www.thinkingbeyond.ai</a>.<h2>References</h2><ul><li><a href="https://cisr.mit.edu/publication/20220101DashboardingWeillWoerner" target="blank" rel="noopener noreferrer">MIT CISR, "Dashboarding Pays Off"</a></li><li><a href="https://www.frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2023.1122200/full" target="blank" rel="noopener noreferrer">Frontiers in Psychology, "Dealing with information overload: a comprehensive review"</a></li><li><a href="https://www.microsoft.com/en-us/worklab/work-trend-index/breaking-down-infinite-workday" target="blank" rel="noopener noreferrer">Microsoft WorkLab, "Breaking down the infinite workday"</a></li><li><a href="https://hbr.org/2024/09/where-data-driven-decision-making-can-go-wrong" target="blank" rel="noopener noreferrer">Harvard Business Review, "Where Data-Driven Decision-Making Can Go Wrong"</a></li><li><a href="https://hbr.org/podcast/2025/03/the-right-way-to-make-data-driven-decisions" target="blank" rel="noopener noreferrer">HBR On Strategy, "The Right Way to Make Data-Driven Decisions"</a></li><li><a href="https://ideas.repec.org/a/eee/epplan/v2y1979i1p67-90.html" target="blank" rel="noopener noreferrer">Donald T. Campbell, "Assessing the impact of planned social change"</a></li><li><a href="https://scholarsarchive.byu.edu/facpub/8217/" target="blank" rel="noopener noreferrer">BYU ScholarsArchive, "Surrogation Fundamentals: Measurement and Cognition"</a></li><li><a href="https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai" target="blank" rel="noopener noreferrer">McKinsey, "The State of AI: Global Survey 2026"</a></li><li><a href="https://www.wavestone.com/en/news/2024-data-and-ai-leadership-executive-survey-41/" target="blank" rel="noopener noreferrer">Wavestone, "2024 Data and AI Leadership Executive Survey"</a></li><li><a href="https://www.linkedin.com/pulse/safe-choice-dangerous-one-zero-buffer-world-maciejewski-zk8je/" target="blank" rel="noopener noreferrer">TBAI #19, "The Safe Choice Is the Dangerous One in a Zero-Buffer World"</a></li></ul><hr><h2>Review package fields (not part of the newsletter)</h2>