Metrics Explained: Meaning, Examples & Why They Matter
Metrics are everywhere, even when we do not consciously think about them. A business tracks revenue, a website owner watches organic traffic, a marketing team measures conversion rates, and a fitness app records steps or heart rate. In each case, a metric turns an activity, result, or condition into information that can be measured and compared. Metrics make complex situations easier to understand because they replace vague impressions with observable data. However, collecting numbers alone does not automatically create useful insight, because every metric needs context and a clear purpose. Understanding how metrics work helps individuals and organizations make smarter, more confident decisions.
Modern organizations now have access to more data than ever before, which makes choosing the right measurements increasingly important. Analytics platforms, customer relationship management systems, financial software, search tools, and operational dashboards can produce hundreds or even thousands of data points. The challenge is therefore not simply finding information but identifying which numbers actually describe progress, performance, efficiency, quality, or risk. Useful metrics connect measurement with a meaningful question, such as whether customers are staying longer or marketing investments are generating profitable growth. Poorly selected measurements can distract teams and encourage decisions based on numbers that have little connection to actual objectives. Good metrics simplify complexity rather than adding more of it.
This guide explains the meaning of metrics, how they differ from KPIs, common examples, major categories, and the reasons they matter in modern decision-making. It also explores how organizations can select meaningful performance indicators, interpret quantitative data correctly, and avoid common measurement mistakes. Whether you work in marketing, finance, technology, operations, human resources, or another field, the basic principles of measurement remain surprisingly similar. Understanding those principles can improve reporting, planning, benchmarking, forecasting, and performance management. More importantly, it helps people ask better questions about what their data actually represents. Once metrics are viewed as decision-making tools rather than isolated numbers, their real value becomes much clearer.
What Are Metrics and What Do They Mean?
A metric is a measurable value used to track, compare, evaluate, or understand a particular activity, process, condition, or result. It usually represents something that can be expressed numerically, such as sales revenue, website visitors, customer satisfaction scores, response time, production output, or employee turnover. Metrics can describe current performance, historical changes, differences between groups, or progress toward an objective. They are commonly used because numbers create a consistent way to evaluate situations that might otherwise feel subjective. A metric does not necessarily tell you whether something is good or bad without additional context. Instead, it provides a defined measurement that can support further analysis and decision-making.
The meaning of a metric depends heavily on what is being measured and why the measurement exists. For example, a website’s monthly traffic is a useful digital marketing metric, but traffic alone cannot reveal whether those visitors become customers. Similarly, increasing production output may appear positive until managers discover that product defects have also increased. This is why useful measurement requires relationships between numbers, business goals, customer outcomes, and operational conditions. Metrics become more informative when compared with previous periods, industry benchmarks, internal targets, competitors, or related indicators. A number viewed alone may create an incomplete picture of performance. Context transforms raw measurement into something that people can interpret and act upon.
Metrics may be expressed as simple counts, percentages, ratios, averages, monetary values, durations, scores, or rates. A company might count the number of new customers acquired during a month, while another measurement calculates customer acquisition cost by dividing spending by the number of customers gained. Percentage metrics are often useful when comparing groups or periods of different sizes because they provide proportional information. Ratios can reveal relationships between two variables, while averages help summarize larger collections of observations. Time-based measurements are especially common in customer service, logistics, manufacturing, and technology operations. The mathematical format should always match the question the organization is attempting to answer.
Some metrics are leading indicators, while others are lagging indicators that describe results after they have occurred. A leading metric can provide an early signal about future performance, such as the number of qualified sales opportunities currently entering a pipeline. A lagging metric records an outcome that has already happened, such as quarterly revenue or customer churn during the previous month. Organizations often need both types because relying entirely on historical results makes it difficult to influence future performance. Leading indicators can encourage earlier intervention, while lagging indicators help determine whether past strategies actually worked. Understanding this distinction improves forecasting and performance monitoring. A balanced measurement system normally includes information about both current activity and completed outcomes.
Metrics can also operate at different levels within an organization, from company-wide measures to highly specific operational indicators. Executives might monitor profitability, revenue growth, cash flow, and customer retention, while a marketing specialist monitors click-through rate, organic traffic, and conversions. A customer service manager may focus on resolution time and satisfaction scores, whereas an IT team tracks uptime, latency, and incident frequency. Each measurement serves a different audience and decision-making purpose. The most valuable metric is therefore not always the most important number for the entire organization. It is the measurement that provides useful information for the particular decision, responsibility, or objective being considered.
Why Metrics Matter for Better Decisions
Metrics matter because they replace assumptions with measurable evidence and make performance easier to evaluate objectively. Without clear measurement, teams may rely heavily on intuition, isolated experiences, or personal opinions when deciding whether something is improving. Data does not eliminate judgment, but it provides a stronger foundation for discussing problems and opportunities. When everyone uses clearly defined performance metrics, conversations can focus on observable results rather than competing impressions. This improves accountability because progress can be monitored using criteria agreed upon in advance. It also makes it easier to identify areas that deserve additional investigation. Measurement therefore supports both everyday operational decisions and larger strategic choices.
Metrics are particularly valuable for identifying trends that may be difficult to notice through individual events. A single decline in weekly sales may mean little, but several months of falling conversion rates can indicate a more meaningful change. Similarly, one customer complaint may be isolated, while a growing complaint rate could reveal a product, service, or communication problem. Trend analysis helps organizations separate temporary fluctuations from patterns that deserve attention. Managers can then investigate possible causes rather than reacting impulsively to every short-term movement. Historical metrics also create baselines against which future results can be compared. Over time, this creates a clearer understanding of normal performance and unusual changes.
Effective metrics also improve goal setting because they make broad ambitions more specific and measurable. A company might say it wants to improve customer loyalty, but that goal becomes easier to manage when supported by retention rate, repeat purchase frequency, or renewal data. Similarly, a marketing department seeking stronger performance can monitor conversion rate, qualified leads, customer acquisition cost, and return on advertising spend. Measurement allows teams to determine how far they are from a desired result and whether current initiatives are helping. It also encourages realistic targets based on actual historical performance rather than arbitrary expectations. Well-designed targets create direction without encouraging teams to manipulate numbers merely to reach a goal.
Another major benefit of metrics is their ability to support resource allocation and prioritization. Organizations have limited money, time, staff, and attention, so leaders must constantly decide where those resources should go. Performance data can reveal which products generate healthy margins, which marketing channels produce valuable customers, or which operational processes create unnecessary delays. Those findings help decision-makers invest in activities that contribute more strongly to strategic objectives. Metrics can also expose underperforming areas that need improvement, redesign, or discontinuation. When resources are assigned based on meaningful evidence, organizations can reduce waste while improving efficiency. This is especially important during periods of rapid growth or economic uncertainty.
Metrics ultimately create value when they lead to better actions rather than simply appearing on reports and dashboards. Organizations sometimes collect large amounts of data because measurement feels sophisticated, even when nobody knows what decisions the information should influence. A useful metric should help someone understand what happened, why it may have happened, or what action deserves consideration next. It should also be reviewed at a frequency appropriate to the activity being measured. Some operational measurements may require daily monitoring, while strategic performance measures may be reviewed monthly or quarterly. Measurement works best when connected to ownership, interpretation, and decision-making. Numbers become valuable only when people know what to do with them.
Metrics vs KPIs: What Is the Difference?
Metrics and key performance indicators are closely related, but the terms do not always mean exactly the same thing. A metric is any defined measurement used to understand an activity, process, or result, while a KPI is a metric considered especially important to a specific objective. In other words, all KPIs are metrics, but not every metric deserves to be treated as a KPI. A business may track hundreds of measurements across departments while selecting only a small number as strategic performance indicators. KPIs usually receive greater attention because they reflect progress toward important business goals. Understanding this distinction prevents dashboards from becoming overloaded with numbers that have equal visual importance but very different strategic value.
Consider an ecommerce business that wants to increase profitable customer growth during the year. Website sessions, product page views, email open rates, and social engagement may all be useful marketing metrics. However, conversion rate, customer acquisition cost, average order value, repeat purchase rate, and contribution margin may be more directly connected to the company’s core objective. Those measurements could therefore become KPIs for the teams responsible for growth. The distinction comes from strategic relevance rather than from the mathematical structure of the metric itself. A measurement can even be a KPI in one organization but a secondary metric in another. Business priorities determine which numbers deserve key status.
KPIs should normally connect directly with outcomes that matter to stakeholders rather than simply recording activity. For example, a sales team could measure the number of phone calls made each day, but call volume does not necessarily indicate successful selling. Qualified opportunities, win rate, sales cycle length, and revenue generated may provide stronger indications of sales effectiveness. Activity metrics still have value because they can help diagnose why results are changing. However, treating every activity measurement as a KPI can encourage employees to optimize quantity rather than meaningful outcomes. Good performance management therefore combines outcome indicators with supporting operational metrics. The supporting data explains performance, while KPIs communicate whether important objectives are being achieved.
The relationship between metrics and KPIs also changes as organizational priorities evolve. A rapidly growing company may initially treat new customer acquisition as its primary focus, making acquisition-related measures central KPIs. Later, the organization may discover that retaining existing customers creates greater profitability, causing churn rate and customer lifetime value to receive more attention. During periods of operational instability, product reliability or service response time might temporarily become executive-level indicators. This flexibility is important because a fixed measurement system can eventually become disconnected from business reality. Organizations should periodically review whether their key indicators still match current priorities. A KPI should remain important because of strategy, not simply because it has always appeared on a dashboard.
Another useful distinction is that KPIs usually include an expected direction, target, benchmark, or threshold. Knowing that customer retention is 82 percent is useful, but the number becomes more meaningful when managers know whether the target is 85 percent or the industry norm is substantially different. Metrics without context can still support analysis, but KPIs are generally expected to show progress toward a defined outcome. This makes them particularly useful for management reporting and strategic reviews. Organizations should therefore avoid calling every available number a KPI merely to make reporting appear more sophisticated. The strongest dashboards highlight a small group of important indicators supported by deeper diagnostic metrics. This hierarchy helps people focus attention where it matters most.
Common Types of Metrics With Practical Examples
Business metrics measure the financial, commercial, and operational health of an organization and are often used by executives and managers. Revenue growth shows how sales change over time, while gross margin provides insight into how much money remains after direct costs. Operating expenses, cash flow, profit margin, and return on investment provide additional perspectives on financial performance. Businesses may also monitor sales pipeline value, customer retention, average revenue per customer, and market share. No single business metric provides a complete picture because profitability, growth, liquidity, and customer behavior can move differently. Combining related measures creates a more balanced view of overall performance and helps leaders understand trade-offs.
Marketing metrics focus on how effectively campaigns and channels attract attention, generate demand, and contribute to customer acquisition. Common digital marketing measurements include impressions, website traffic, engagement rate, click-through rate, cost per click, conversion rate, and cost per acquisition. Search engine optimization teams may monitor organic clicks, search visibility, keyword rankings, conversions, and revenue attributed to organic search. Email marketers can measure open rates, click rates, unsubscribe rates, and campaign conversions, while paid media teams often examine return on ad spend. Strong marketing measurement connects activity with meaningful business outcomes. Large traffic numbers may look impressive, but qualified leads and profitable customers usually provide more strategic value.
Customer metrics help organizations understand satisfaction, loyalty, retention, purchasing behavior, and the overall customer experience. Customer retention rate indicates how successfully a company keeps customers over a specific period, while churn rate measures how many customers leave. Customer lifetime value estimates the economic value a customer may generate throughout the relationship with a business. Companies can also monitor repeat purchase rate, support ticket frequency, refund rate, satisfaction scores, and recommendation-related survey results. These measurements can reveal whether growth is sustainable rather than being driven entirely by constant customer acquisition. When retention improves, companies often benefit from stronger recurring revenue and more predictable demand. Customer data therefore deserves attention alongside traditional sales measurements.
Operational metrics measure efficiency, productivity, quality, capacity, reliability, and process performance inside an organization. A manufacturer might monitor production volume, defect rate, equipment downtime, inventory turnover, and cost per unit. A logistics company may track delivery time, order accuracy, shipping cost, vehicle utilization, and on-time delivery percentage. Customer support operations frequently measure first response time, average handling time, resolution rate, ticket backlog, and service level performance. Technology teams may track system uptime, application latency, error rates, deployment frequency, and incident recovery time. These metrics help managers identify bottlenecks and recurring problems. They are particularly useful when teams need to improve processes without sacrificing quality or customer satisfaction.
Human resources and employee metrics provide insight into workforce stability, hiring efficiency, development, engagement, and organizational capacity. Common HR measurements include employee turnover, absenteeism, time to hire, cost per hire, offer acceptance rate, internal promotion rate, and employee engagement scores. Organizations may also track training participation, performance distribution, workforce productivity, retention among high-performing employees, and representation across relevant workforce categories. These figures should be interpreted carefully because employees cannot be understood through numbers alone. Qualitative feedback, manager observations, and workplace context remain important. When used responsibly, workforce analytics can reveal trends that help organizations improve hiring processes, employee experience, succession planning, and long-term workforce strategy.
How Metrics Are Calculated and Interpreted
Many metrics are calculated using straightforward formulas, but understanding the variables behind the calculation is more important than memorizing equations. A conversion rate, for example, is commonly calculated by dividing the number of desired actions by the relevant number of opportunities and multiplying by one hundred. If 100 people complete a purchase after 2,000 qualified website visits, the conversion rate would be five percent. However, interpretation depends on how visits and conversions are defined. Different analytics platforms or teams may count sessions, users, transactions, and conversion events differently. Clear definitions ensure that everyone interpreting the metric is actually discussing the same measurement.
Rate metrics are especially valuable because they allow fairer comparisons between groups or periods with different volumes. Imagine that one support team receives 1,000 tickets and resolves 900, while another receives only 200 and resolves 190. Comparing resolution counts alone would make the first team appear more productive, but their resolution rates are 90 percent and 95 percent respectively. Neither figure automatically proves which team performs better because ticket complexity and staffing may differ. Nevertheless, rates provide a standardized perspective that raw counts cannot always offer. This is why percentages and ratios appear frequently in business analytics. Normalizing data helps people compare performance more meaningfully across changing conditions.
Averages are another common measurement method, but they can sometimes hide important differences inside the underlying data. Suppose the average customer order value is $100, yet some customers spend $20 while a small number spend several hundred dollars. The average provides a useful summary, but it does not explain the distribution of spending. Median values, ranges, percentiles, segments, and frequency distributions can provide additional context when the data contains significant variation. Similar issues arise with average salary, average response time, average session duration, and average transaction value. Analysts should therefore understand how summary statistics can simplify complex information. A useful interpretation often requires looking beyond a single calculated figure.
Time comparisons are among the most common methods for interpreting performance metrics. Organizations may compare this week with last week, this month with the previous month, or the current year with the same period one year earlier. Year-over-year comparisons can be especially useful for seasonal businesses because they reduce distortions caused by predictable seasonal demand. However, unusual events, changes in tracking systems, product launches, pricing adjustments, or market disruptions can still make comparisons misleading. Analysts should examine whether the underlying conditions are reasonably comparable before drawing conclusions. A percentage increase can look dramatic when the starting value was extremely small. Good interpretation therefore combines numerical change with business context.
Segmentation can make metrics considerably more useful by revealing differences hidden inside an overall result. A company may have a healthy average customer retention rate while discovering that customers acquired through one particular marketing channel leave much faster. Website conversion rates may also vary by device, traffic source, location, product category, or customer type. Sales performance can differ by territory, account size, industry, or representative, while operational outcomes may vary by facility or shift. Segmenting data helps analysts discover patterns that aggregated measurements can conceal. However, excessive segmentation can create noise when sample sizes become too small. The goal is to divide data in ways that support useful decisions rather than generating endless combinations.
How to Choose Metrics That Actually Matter
Choosing useful metrics begins with identifying the decision or objective that measurement is supposed to support. Teams often make the mistake of opening an analytics platform, viewing whatever data is readily available, and then deciding that those numbers must be important. A better approach starts with a question such as whether customer retention is improving, whether acquisition is profitable, or whether service quality meets expectations. Once the question is clear, teams can identify measurements that provide relevant evidence. This objective-first approach reduces unnecessary reporting and keeps attention focused on results that matter. Measurement should follow strategy instead of allowing available data to define strategy.
Good metrics should be clearly defined so that different people calculate and interpret them consistently. If a company tracks an active customer, everyone should understand exactly what activity qualifies someone as active and during what period. Similar clarity is needed for terms such as qualified lead, conversion, retained customer, resolved ticket, employee turnover, and recurring revenue. Ambiguous definitions create reporting disagreements even when everyone is working from the same underlying data. A metric dictionary or documented measurement framework can help prevent this problem. Definitions should include the calculation method, data source, reporting frequency, and responsible owner. Consistency becomes increasingly important as organizations grow and more teams depend on shared analytics.
Useful metrics should also be actionable, meaning that people can reasonably influence the factors contributing to the result. An organization can track broad economic conditions, but individual teams usually cannot control them directly. By contrast, conversion rate, response time, product quality, and customer retention may be affected by specific processes and decisions. Actionability does not mean every measurement must be entirely controllable because external forces influence most business outcomes. Instead, the metric should provide information that helps someone decide what to investigate or improve. Teams should know what kinds of actions might follow from a significant change. Data that never affects a decision may not deserve prominent reporting space.
Balanced measurement is equally important because optimizing one metric can sometimes damage another. A customer service team could reduce average handling time by rushing conversations, but customer satisfaction and first-contact resolution might decline. A manufacturer could increase production volume while producing more defective items, and a marketing team could generate cheaper leads that convert poorly into paying customers. These examples show why single-metric optimization can create unintended consequences. Organizations should pair efficiency metrics with quality, profitability, or customer outcome measures whenever appropriate. The purpose is not to create dozens of KPIs but to protect against narrow interpretations. Balanced scorecards help teams understand both the benefits and potential costs of performance improvements.
Metrics should finally be reviewed and updated as products, markets, business models, and organizational priorities change. A startup entering a new market may care intensely about adoption and user growth, while a mature organization may focus more heavily on profitability and retention. A metric that once guided an important decision can gradually become less relevant as conditions evolve. Teams should therefore periodically ask whether each important measurement still helps explain progress toward current goals. Removing outdated metrics is just as important as adding new ones because crowded dashboards reduce attention. Effective measurement systems remain focused, understandable, and adaptable. The objective is not maximum data collection but maximum decision value.
Common Mistakes When Using Metrics
One of the most common measurement mistakes is focusing on vanity metrics that look impressive but reveal little about meaningful performance. Social media followers, page views, app downloads, and email subscribers can be useful measurements in the right context, yet large numbers do not necessarily translate into revenue, retention, engagement, or customer value. A website could double its traffic while generating exactly the same number of qualified leads. Similarly, an application could attract thousands of downloads that never become active users. Vanity metrics become problematic when they are presented as evidence of success without a connection to business outcomes. Strong reporting explains what a number means rather than relying on its size to create an impression.
Another mistake is assuming that correlation proves causation simply because two metrics move together. Marketing spending and revenue might increase during the same period, but that does not automatically prove the campaign caused all of the growth. Seasonal demand, pricing changes, competitor activity, product improvements, or broader economic conditions may also contribute. Similar errors occur when teams attribute churn, employee turnover, website traffic changes, or productivity shifts to a single factor without sufficient evidence. Metrics can highlight relationships that deserve investigation, but they do not always explain why those relationships exist. Careful analysis considers alternative explanations. Experiments, segmentation, controlled comparisons, and qualitative research can help strengthen causal understanding.
Using inconsistent measurement periods can also produce misleading conclusions. Comparing a 31-day month with a 28-day month based only on total activity may exaggerate differences that largely result from the number of days. Weekend patterns, holidays, seasonal events, campaigns, and promotional periods can create similar distortions. Teams should choose comparison periods that match the question being investigated and normalize figures when appropriate. Daily averages, weekly rates, or year-over-year comparisons may sometimes provide a clearer view than raw totals. Analysts should also document changes in tracking systems because new definitions can create artificial jumps or declines. Consistency makes historical trend analysis much more reliable.
Ignoring data quality is another serious problem because a precise-looking dashboard can still contain inaccurate information. Tracking scripts can fail, duplicate events may inflate counts, CRM records can become incomplete, and integrations may stop transferring data correctly. Human entry errors can also affect sales, finance, HR, and operational databases. Organizations should therefore validate important metrics rather than assuming every number produced by software is automatically trustworthy. Sudden unexplained changes deserve investigation, especially when no corresponding business event occurred. Data governance, validation rules, ownership, and routine quality checks help reduce reporting errors. Reliable decisions require reliable inputs, regardless of how attractive the visualization appears.
Perhaps the most damaging mistake is allowing metrics to become targets that employees learn to manipulate instead of measures that help improve outcomes. If a team is rewarded entirely for the number of calls completed, employees may prioritize short calls even when customers need more assistance. If marketers are evaluated solely on lead volume, they may pursue low-quality leads that sales representatives cannot convert. Measurement systems influence behavior, so incentive design deserves careful attention. Managers should ask what behavior a metric encourages and whether that behavior supports the intended objective. Combining quantitative indicators with judgment and qualitative context can reduce unhealthy optimization. Metrics should guide better performance rather than becoming the performance itself.
How to Build a Better Metrics and Reporting System
A strong measurement system begins with a small number of clearly defined objectives rather than a massive dashboard containing every available data point. Leadership should determine which outcomes are strategically important and then identify the measurements that provide evidence about those outcomes. Supporting teams can add diagnostic metrics that help explain why the primary indicators move. This creates a hierarchy in which executives see high-level performance while specialists retain access to deeper operational data. The structure helps prevent information overload without eliminating valuable detail. Good reporting therefore separates strategic KPIs from supporting metrics. Everyone can then focus on the information most relevant to their responsibilities.
Ownership is another important part of effective measurement because every major metric should have someone responsible for understanding and maintaining it. The owner does not necessarily control the entire result, but they should understand the definition, source, reporting process, and factors that influence the measurement. Without ownership, dashboards can accumulate outdated numbers that nobody investigates when something changes unexpectedly. Clear responsibility also makes it easier to resolve disagreements about calculations or data quality. Teams should know who can explain each important number and where the underlying data originates. This creates accountability without turning metrics into tools for assigning blame. Measurement works better when responsibility encourages learning and improvement.
Reporting frequency should match the speed at which a metric changes and the speed at which decisions can realistically be made. Website uptime may require continuous monitoring because technical incidents demand rapid action, while strategic customer lifetime value may not need daily review. Checking slow-moving metrics too frequently can encourage overreaction to random fluctuations. On the other hand, reviewing rapidly changing operational information only once per quarter may allow problems to continue unnecessarily. Organizations should choose daily, weekly, monthly, quarterly, or real-time reporting based on decision needs. A useful dashboard is therefore not merely a collection of numbers but a carefully designed information system. Timing influences how effectively people respond to what they see.
Visualization also plays a major role in how metrics are interpreted and communicated. Trend lines can show whether performance is improving or deteriorating over time, while comparison charts can highlight differences among products, regions, or customer segments. Scorecards can display current values alongside targets, and tables may provide detailed information when exact numbers matter. Visual design should clarify patterns rather than decorate the data with unnecessary complexity. Labels, units, reporting periods, and definitions should be immediately understandable to the intended audience. Color or alerts can highlight meaningful exceptions, but excessive visual signals can reduce their impact. Simple dashboards are often more actionable than complicated ones filled with dozens of widgets.
The strongest reporting systems encourage questions rather than pretending that metrics provide complete answers automatically. When a KPI changes, teams should ask what contributed to the movement, whether the change is statistically or commercially meaningful, and what additional evidence is needed. Quantitative data can then be combined with customer interviews, employee feedback, market knowledge, operational observations, and experimentation. This prevents organizations from treating numbers as substitutes for understanding. Metrics are tools that help people observe patterns and test ideas, but interpretation still requires judgment. A mature measurement culture values curiosity as much as reporting accuracy. The goal is ultimately to turn information into better decisions, stronger performance, and continuous learning.
Frequently Asked Questions About Metrics
What is a metric in simple terms?
A metric is a measurable value used to track or evaluate something, such as revenue, website traffic, response time, customer retention, or sales growth. It helps turn an activity or result into data that can be compared and analyzed.
What is an example of a metric?
Conversion rate is a common metric that shows the percentage of people who complete a desired action, such as making a purchase or submitting a form. Other examples include profit margin, employee turnover, customer acquisition cost, website traffic, and delivery time.
What is the difference between a metric and a KPI?
A metric is any useful measurement, whereas a KPI is a metric considered especially important for evaluating progress toward a specific goal. Every KPI is therefore a metric, but many supporting metrics are not important enough to become key performance indicators.
Why are metrics important in business?
Metrics help businesses understand performance, identify trends, compare results, allocate resources, and make decisions based on measurable evidence. When selected carefully, they also reveal whether strategies and operational improvements are producing the intended outcomes.
How do you choose the right metrics?
Start with the objective or decision you need to support and then select measurements that directly provide relevant information about that goal. Strong metrics are clearly defined, reliable, actionable, understandable, and balanced with other measurements that provide necessary context.
