Most businesses are drowning in metrics. Dashboards everywhere, KPIs for everything, quarterly reviews packed with charts. But ask what actually drives business outcomes and you’ll often find critical gaps in what’s being measured.

The patterns are consistent across industries and company sizes. We measure what’s easy to measure, what makes us look good, or what doesn’t threaten existing power structures. We avoid measuring things that might reveal uncomfortable truths.

The Vanity Metrics Problem

Revenue and user growth get tracked obsessively. They’re good numbers to report. Investors like them. They trend upward (hopefully), which creates positive narratives in all-hands meetings.

Customer acquisition cost and customer lifetime value should be tracked with equal rigor, but often aren’t. CAC is rising and LTV is falling at many companies, but those trends don’t get the same visibility as top-line growth metrics.

Why? Because if CAC is increasing faster than LTV, it raises questions about business model sustainability. Those questions are uncomfortable. They might imply the current growth strategy isn’t working. Easier to focus on growth numbers and hope the unit economics eventually sort themselves out.

Employee Retention Nobody Wants to Confront

Most companies track headcount and hiring velocity. Fewer track retention with the same intensity, and fewer still try to understand why people leave.

Exit interviews happen, but they’re often perfunctory. The departing employee doesn’t want to burn bridges, so they give diplomatic answers. “Pursuing new opportunities” reveals nothing about whether they left due to poor management, lack of growth, or compensation issues.

What would actually help is measuring retention by manager, by team, and by cohort. If one department has 40% annual turnover while others have 10%, that’s diagnostic. But making those numbers visible might create awkward questions about specific leaders.

Some companies actively resist this level of measurement because they know what it would show. Better to treat retention as a company-wide challenge rather than identify which managers are driving people out.

Customer Complaints That Disappear

Support ticket volume gets measured. Resolution time gets measured. Customer satisfaction scores get measured. But the actual content of complaints often doesn’t get systematically analyzed.

I’ve seen companies where the support team knows exactly what the top customer pain points are, but that knowledge never reaches product or leadership because there’s no formal mechanism to aggregate and elevate complaint themes.

Measuring complaint categories would mean acknowledging product gaps or design failures. It might mean admitting a recently launched feature is confusing users. It might reveal that the “simplified” pricing structure actually made things more complicated.

Leadership gets reports showing high customer satisfaction scores (because most people who interact with support are ultimately satisfied with the resolution) and doesn’t see the pattern of recurring problems in the product itself.

The Real Cost of Technical Debt

Engineering teams know their codebases are accumulating technical debt. They know which systems are fragile, which features require excessive maintenance, which architectural decisions are causing problems.

But technical debt rarely gets measured in terms executives understand. There’s no line item for “hours spent working around our database design problems” or “features we didn’t ship because the frontend framework is outdated.”

Creating those measurements would mean quantifying the cost of past shortcuts. It might mean admitting that the aggressive shipping schedule from two years ago created inefficiencies that are still compounding. That’s uncomfortable for the leaders who pushed that schedule.

Better to keep technical debt as a vague concept that engineering complains about rather than a measured cost center that demands attention.

Sales Pipeline Realism

Sales organizations measure pipeline extensively—number of deals, deal sizes, probability-weighted forecasts. But they often resist measuring forecast accuracy against actual outcomes.

How often do deals marked “90% likely to close this quarter” actually close on time? For many sales organizations, that number would be embarrassing. The pipeline systematically overstates near-term revenue.

Everyone knows this is happening, but making it explicit would force difficult conversations about sales methodology, forecasting discipline, or whether the sales team is being honest about deal status.

Process Efficiency vs Process Existence

Companies love to measure whether processes are being followed. Did we complete the quarterly review? Did all employees finish compliance training? Did we hold the required number of candidate interviews?

They measure far less often whether those processes actually achieve their intended outcomes. Does the quarterly review process lead to better decision-making? Does compliance training reduce actual compliance violations? Does the interview process improve hire quality?

Measuring outcomes rather than activity might reveal that expensive, time-consuming processes aren’t providing value. That would mean either fixing them or eliminating them. Both options require effort and potentially admitting that established practices aren’t working.

The Metrics That Would Actually Help

What should companies measure that they currently don’t?

Time from decision to implementation. How long does it take from “we should do X” to actually doing X? Long delays indicate bureaucracy problems, resource constraints, or lack of follow-through.

Failed initiative rate. What percentage of launched projects achieve their stated goals? If most initiatives fail or fizzle out, that suggests problems with planning, execution, or goal-setting.

Key person dependencies. Which individuals are critical to too many processes? If one person leaving would cripple multiple workflows, that’s organizational fragility worth measuring and addressing.

Decision reversal frequency. How often do decisions get made and then unmade weeks later? High reversal rates might indicate poor decision processes or lack of leadership alignment.

All of these metrics would reveal organizational problems. That’s exactly why they don’t get measured.

The Political Dimension

Metrics aren’t politically neutral. What gets measured affects what gets attention and resources. Whose performance gets highlighted and whose gets obscured.

Proposing new metrics is implicitly suggesting existing metrics are incomplete or misleading. That can be seen as criticism of whoever chose the existing metrics.

Measurement systems entrench power structures. If marketing success is measured by lead volume rather than lead quality, marketing looks good while sales struggles with unqualified leads. Changing the metric threatens marketing’s perceived performance.

Creating Measurement Systems That Work

The companies that measure well create safe systems for surfacing uncomfortable truths. Leadership actively seeks out metrics that might reveal problems because they want to fix problems before they become crises.

They separate measurement from blame. High employee turnover in one department isn’t automatically that leader’s fault—it’s diagnostic information that warrants investigation. Maybe compensation is uncompetitive. Maybe workload is unsustainable. Maybe the manager does need coaching. But the metric comes first, judgment comes after investigation.

They regularly audit what’s being measured against what actually drives business outcomes. If you’re tracking 20 metrics but only three actually correlate with long-term success, focus on those three.

They’re willing to measure themselves honestly even when the numbers are bad. Especially when the numbers are bad, because that’s when measurement provides the most value.

Start With One Honest Metric

If you’re in a position to influence measurement in your organization, pick one thing that matters but isn’t currently measured. Something where you suspect the truth would be uncomfortable but useful.

Measure it for yourself first. Understand what the data shows before deciding whether to make it visible more broadly. Sometimes the uncomfortable truth is actually not that bad, and resistance was based on vague fear rather than actual problems.

If the data does reveal issues, figure out whether the organization is ready to act on that information. There’s no point measuring something if leadership isn’t willing to respond to what the measurement shows.

But don’t let fear of uncomfortable truths prevent measurement entirely. The problems you’re not measuring are still there. They’re just festering invisibly instead of being addressed. Measurement at least gives you the option to fix things.

Most businesses have the capability to measure what matters. They just need the courage to actually do it.