You may find yourself in front of a dashboard full of numbers that nobody looks at. You may discover that two departments measure the same thing in different ways. You may notice that the indicators change every time the monthly meeting changes.
\nThese are recurring symptoms, reported by the academic literature on small and medium-sized companies in Europe as structural patterns, not isolated incidents [5].
\nPerformance measurement is the system a company uses to decide what to observe in order to understand whether it is heading where it wants to go. It isn't a list of indicators; it's the map that links indicators to strategy. Structured management practices — monitoring indicators, setting goals, incentives — are associated with higher productivity levels in the survey conducted by the Bank of Italy on about 3,200 Italian companies with at least twenty employees [7].
\nThis article walks through building a coherent measurement system: from the operational definition to the strategic cascade, through the reference models (Balanced Scorecard, OKR, Performance Prism), all the way to the mistakes that strip meaning even from the best-designed architectures.
\nIn your company, is there a map that explains why certain things are measured and others aren't? If the answer doesn't come immediately, the problem isn't the number of indicators — it's the absence of a system that justifies the measurement choices.
\nThe word \"measurement\" brings to mind sophisticated dashboards and business intelligence software. The reality in many Italian companies is different: an Excel spreadsheet updated by hand, a monthly report nobody reads, a quarterly meeting where the numbers are discussed without a clear logic. This section proposes an operational definition of a measurement system and draws its boundaries with respect to related concepts. The distinction matters: those who confuse the system with the single indicator tend to collect numbers without a map that relates them to one another [5].
\nOperational definition. A performance measurement system is the set of choices (which areas to observe, which indicators to use, how often, with what responsibility, in service of which decisions) that allows the company to assess how it is doing against its stated goals. It isn't software, it isn't a report, it isn't a set of KPIs: it's the logic that links them.
\nThe four boundaries to clarify.
\nMeasurement vs evaluation. Measurement is objective and quantitative — it records what happened through indicators defined in advance. Evaluation is interpretive — it assigns a judgment to what was measured, in light of goals, context and priorities. The same data point can be positive or negative depending on how it is evaluated.
\nMeasurement vs KPIs. A KPI is a single indicator. Measurement is the system that decides which indicators to choose, how to build them and how to read them together. Confusing the system with the indicator leads to collecting KPIs without a map that relates them to one another. For detail on individual indicators, the cluster on business KPIs is the reference.
\nMeasurement vs reporting. Measurement concerns what is measured; reporting concerns how what was measured is communicated. A good report on a poorly designed measurement system is still a useless report. For the process that uses measurement results to produce decisions, the cluster on management control is the complementary reference.
\nMeasurement vs benchmarking. Measurement is internal — it compares the company with itself over time. Benchmarking is comparative — it compares the company with other organizations in the industry or with reference standards. The two processes coexist but answer different questions.
\nFor the organizational framework in which the measurement system fits as a component of business systemization, the area pillar is the point of reference.
\nWhat is the first critical question the measurement system should answer? If the first thing that comes to mind is a tool — not a question — the sequence is reversed.
\nThe systematic review by Bititci and colleagues of more than 6,000 academic articles identifies a recurring mistake in measurement systems: the choice of tools comes before the choice of goals [4]. Companies implement dashboards before clarifying which questions the system should help answer. This section proposes a three-step method for deciding what to measure: start from the critical business questions, derive the performance areas to observe, and choose indicators only as the last step. The principle is simple: tools follow questions, not the other way around.
\nThe three-step method.
\nStep 1 — Identify the critical business questions. Critical questions are the ones the business owner can't answer today, but whose answer would change operational decisions. Examples: \"Which product line is most profitable?\", \"Which customers generate the best margin?\", \"How long does it take us from receiving an order to delivery?\", \"Which stage of the process has the highest error rate?\".
\nStep 2 — Derive the performance areas to observe. Each critical question points to a performance area to monitor. The question \"which line is most profitable?\" points to the area of profitability by product line. The question \"how long does it take us from receipt to delivery?\" points to the area of throughput times. Performance areas are the intermediate level between strategy and indicators.
\nStep 3 — Choose the indicators. Only after identifying the questions and the areas do you choose the indicators. For each area, identify 1-3 measurable indicators, with a measurement frequency and an owner. Indicators multiply almost always because step 1 (the critical questions) was skipped.
\nOperational framework.
\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n| Critical question | Performance area | Indicator | Frequency |
|---|---|---|---|
| Which line is most profitable? | Profitability by line | Contribution margin by line | Monthly |
| How long from receipt to delivery? | Throughput times | Average lead time by type | Weekly |
| Which customer generates the best margin? | Profitability by customer | Margin per top-20 customer | Quarterly |
scroll the table →
\nHow many of the indicators you measure today can be directly linked to an explicit strategic goal? When the link isn't immediate, the indicator is probably measuring something that doesn't help you decide.
\nWhen Andy Grove introduced the OKR system at Intel in the late 1970s, the rule was explicit: every lower-level objective had to be traceable to a higher-level objective [2]. Twenty years later, John Doerr brought the same principle to Google. The cascade principle — operational indicators descend from strategic goals, never the other way around — is at the heart of any coherent measurement system. This section describes how to structure the cascade on three levels and how to keep it from breaking at the intermediate steps.
\nThe three-level cascade.
\nLevel 1 — Strategic goals. What do you want to achieve over the next 3-5 years? Strategic goals are few (3-5), ambitious, and rarely expressed as precise numbers. \"Becoming the go-to supplier for B2B customers in northeastern Italy\" is a strategic goal. \"Increasing revenue by 15%\" can be a strategic goal if it is connected to a vision, not if it is an arbitrary target.
\nLevel 2 — Performance areas and process goals. Each strategic goal translates into 3-5 performance areas and into process goals measurable over 12 months. The goal \"becoming the go-to supplier\" translates into areas such as \"service quality\", \"response times\", \"contract renewal rate\". Each area has a measurable goal for the current year.
\nLevel 3 — Operational indicators and activities. Each performance area has 1-3 operational indicators measured weekly or monthly. Operational indicators feed day-to-day decisions.
\nHow to keep the cascade from breaking. The most frequent breaking point is between level 1 and level 2: strategic goals stay on paper in the annual planning meetings and are never translated into performance areas and process goals. The cause is almost always the absence of an explicit moment (a dedicated working session) in which the translation is done. For the cluster on business KPIs, which goes into detail on the operational indicators to include in the cascade, the reference is available. For the connection with business process mapping as the basis for level 2 indicators, the dedicated cluster is complementary.
\nWhich model best suits a company with fast decision cycles and significant external stakeholders? The answer depends on three variables that are often not considered before the choice.
\nIn 1992 Robert Kaplan and David Norton proposed in the Harvard Business Review a model that broke with tradition: measuring performance not only through financial indicators, but through four integrated perspectives [1]. The Balanced Scorecard is today the most widespread measurement framework. In the following years alternative models emerged — Andy Grove and John Doerr's OKRs [2], the Performance Prism from the Cranfield School of Management [3] — each with a different center of gravity. This section compares the three models on an operational level: what they measure, in which context they work best, and what the adoption curve looks like for a smaller company.
\nBalanced Scorecard (Kaplan-Norton, 1992) [1].\nFocus: linking strategy to measurement through four balanced perspectives (financial, customer, internal processes, learning and growth). Strengths for smaller companies: it forces you to reason about several dimensions at once, not just the financial numbers. Limits for smaller companies: building it takes time and organizational commitment; the four perspectives can feel over-structured for companies with fewer than 20-30 employees.
\nOKR — Objectives and Key Results (Grove-Doerr) [2].\nFocus: quarterly cycles of ambitious objectives with 2-4 measurable key results. Strengths for smaller companies: a fast cadence (quarterly instead of annual), a simple formulation (objective + key results). Limits for smaller companies: it requires a quarterly review discipline that degrades quickly if it isn't maintained. It doesn't include a structured logic for non-financial perspectives.
\nPerformance Prism (Cranfield School of Management, Neely et al.) [3].\nFocus: starting from the satisfaction and contribution of stakeholders (not only shareholders) to derive strategy, processes and capabilities. Strengths for companies with a broad stakeholder base: supply chain, employees, local community, banks — all included as relevant perspectives. Limits for smaller companies: less widespread as an operational tool; it requires a preliminary stakeholder review that can be complex.
\nWhich model for which context.
\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n| Variable | Balanced Scorecard | OKR | Performance Prism |
|---|---|---|---|
| Decision cycle | Annual-semiannual | Quarterly | Annual |
| Organizational structure | Mid-size to larger company | Small to mid-size company | Company with a broad stakeholder base |
| Stakeholders | Mainly financial and customers | Mainly the internal team | Multi-stakeholder |
| Implementation complexity | Medium-high | Low | High |
scroll the table →
\nFor a company implementing a measurement system for the first time, the operational recommendation is to start with a simplified version of OKRs (objectives + 2-3 key results per quarter) and introduce the Balanced Scorecard structure gradually, one perspective at a time, over 12-18 months.
\nWhat really changes when you move from an Excel spreadsheet to an integrated dashboard? The tool doesn't add intelligence to the system — it reveals it or exposes it, depending on how structured it already was.
\nIn Italy, 48.8% of small and medium-sized companies use ERP software, but only 21.1% have adopted CRM systems [6]. Data analysis tools stop at 41.9% among Italian small and medium-sized companies, compared with 83.6% of large companies: a gap of more than forty percentage points [6]. The good news is that the maturity of the measurement system doesn't depend on the tool. The bad news is that without a clear system behind it, even the most sophisticated platform produces useless dashboards.
\nThree levels of tooling.
\nBasic level — Spreadsheets. Suitable for companies with up to 20-30 employees, with a simple measurement system (3-5 indicators, monthly frequency). A structured spreadsheet with the goals-indicators cascade and a monthly comparison between target and actuals is enough for the early stages. The main limit is dependence on the person who manages it: when that person changes, the system stops.
\nIntegrated level — Reporting modules in management software. Suitable for companies with 20 to 60 employees that already have an ERP or integrated management software. The reporting module reads data directly from accounting and sales, reducing the risk of transcription errors. The limit is dependence on the quality of the data entered into the management software.
\nAdvanced level — Dedicated performance management platforms. Suitable for companies with more than 60-80 employees, several divisions, cost centers and the need to consolidate different data sources. Implementation complexity is significant: without a dedicated internal owner, the risk of abandonment within 12-18 months is high. For the connection with the management control process that uses tools to produce operational decisions, the dedicated cluster is the reference.
\nSelection criteria. Four variables guide the choice: the number of indicators to monitor (the higher it is, the more a dedicated tool is justified), the required update frequency (daily requires automation, monthly can be manual), the number of users who read the reports (beyond 5-7, a shared tool is preferable), and integration with accounting (data already in the management software reduces manual work).
\nOf the five mistakes described, which is the most costly when it shows up silently? It isn't indicator overload: it's the wrong reading frequency, because it creates late decisions that look timely.
\nThe review by Bititci and colleagues of performance measurement systems identifies a series of failure patterns that repeat with surprising consistency, regardless of industry and size [4]. The same picture emerges from the literature specific to small and medium-sized companies [5]. This section describes the five most recurrent mistakes, with the warning sign to watch for and a self-diagnosis question you can ask in the next 48 hours.
\nMistake 1 — Indicator overload.\nDescription: the system monitors 20, 30, 50 indicators. Nobody can read them all, and meetings get longer without producing decisions.\nWarning sign: in meetings, fewer than a third of the indicators in the report are discussed.\nSelf-diagnosis question: which indicators, had they been at zero this month, would have changed anything in operational decisions? Only those should be kept.
\nMistake 2 — Misalignment between indicators and strategy.\nDescription: you measure things that were important two years ago, but the strategy has changed. The indicators reflect the past, not the future.\nWarning sign: the indicators in the report have changed by less than 20% in the last 18 months, even though the business has evolved.\nSelf-diagnosis question: can every indicator in the system be linked to an explicit strategic goal for the current year?
\nMistake 3 — Financial indicators only.\nDescription: the system measures revenue, costs and margins, but nothing on service quality, process times or customer satisfaction. Problems are discovered only when they show up in the financial numbers — too late to intervene upstream.\nWarning sign: the monthly report contains only income statement items and no operational or process indicators.\nSelf-diagnosis question: is there at least one non-financial indicator in the system, measured at least monthly?
\nMistake 4 — The wrong reading frequency.\nDescription: indicators are read quarterly when the decision requires monthly updates, or vice versa. The measurement frequency doesn't match the frequency of the decisions it is meant to support.\nWarning sign: in the last three report review meetings, at least once someone said \"but this data is already old\".\nSelf-diagnosis question: is every indicator read with the same frequency at which the decision associated with it is made?
\nMistake 5 — No owner for each indicator.\nDescription: the system produces numbers, but it isn't clear who is responsible for interpreting them and taking action when they deviate from the target.\nWarning sign: in meetings, the data are discussed without anyone saying \"I'll do X by Y\".\nSelf-diagnosis question: does every indicator in the system have a named person responsible for reading it and acting on it?
\nSources [1], [2], [3]. Kaplan-Norton, Doerr and Neely et al. are established management and academic references, not quantified empirical studies. Statements about the characteristics of the models derive from these texts. Statements about failure patterns rely on [4] and [5] (peer-reviewed).
\nSource [6] ISTAT. The adoption rates cited (48.8% ERP, 21.1% CRM, 41.9% data analysis tools for small and medium-sized companies; 85.9%, 56.5% and 83.6% for large companies) are taken verbatim from the ISTAT press release on Italian companies. The survey concerns the software companies have, not the quality of the measurement system that uses it: the link between the two dimensions is an operational inference.
\nTransferability. The review by Bititci et al. [4] is based on international studies, mostly of medium-to-large companies. Garengo et al. [5] focuses specifically on small and medium-sized companies in Europe, but the sample covers contexts other than Italy. The operational guidance has general validity; calibrating it to a specific national context requires case-by-case verification.
\nHow many KPIs does a 20-person company need?\nThere is no universal number, but there is a practical rule: 3 to 7 indicators per organizational level. A business owner with 20 employees can manage 5-7 strategic indicators; each function manager monitors 3-5 in their own area. Beyond this threshold, the system produces data without producing decisions.
\nHow often should KPIs be read?\nThe frequency depends on the type of indicator: operational indicators (lead time, error rate) may need weekly reading; strategic indicators (profitability by line, customer renewal rate) are read monthly or quarterly. The rule is that the reading frequency matches the frequency of the decision the indicator supports.
\nBalanced Scorecard or OKR for a smaller company?\nIf you are starting from scratch, OKRs are more accessible: a simple formulation, a quarterly cycle, no specific training required. The Balanced Scorecard is more structured and better suited to companies that already have a reporting system and want to link it to strategy systematically.
\nA performance measurement system is not a list of KPIs: it is the map that links strategic goals to operational indicators through a three-level cascade (strategy → performance areas → indicators). You build it starting from the critical business questions, not from the tools. You equip it with the minimum level of tooling needed for your organizational maturity. You keep it alive by assigning an owner to each indicator and checking quarterly that the cascade is still aligned with the strategy.
\nThe five mistakes that strip it of meaning (overload, misalignment, financial indicators only, wrong frequency, no owners) are avoided with a single practice: defining the critical questions before choosing the indicators.
\nBuilding a performance measurement system doesn't mean choosing indicators: it means deciding which questions the company wants to ask itself and in what order. The difference between a useful dashboard and a decorative one isn't in the tool, but in the map that precedes the indicators — the cascade that links strategic goals, performance areas, operational indicators and reading tools.
\nTo go deeper into the individual building blocks of the system, it's worth reading the cluster on business KPIs, which covers the choice of individual indicators, and the one on management control, which describes the ongoing process that translates measurement into operational decisions. To place the measurement system within the broader architecture of running a business, the reference is the pillar on business management.
\nA company that truly measures — not for compliance, but to decide — is a company that recognizes trends before they become problems, distributes decisions because everyone sees what concerns them, and allocates resources based on evidence rather than beliefs. The spread of structured management practices — measurement included — is uneven across Italy, and it is one of the dimensions on which the Bank of Italy observes, together with lower productivity, the companies that adopt them least [7]. Seen this way, the measurement system stops being a technical exercise and becomes a component of overall economic health.
\n[1] Kaplan, R. S. and Norton, D. P., \"The Balanced Scorecard — Measures That Drive Performance\", Harvard Business Review, 70(1), January-February 1992, pp. 71-79 (the article that introduces the model); by the same authors, \"The Balanced Scorecard: Translating Strategy into Action\", Harvard Business School Press, 1996. Available at: https://hbswk.hbs.edu/item/the-balanced-scorecard-translating-strategy-into-action
\n[2] Doerr, J., \"Measure What Matters — How Google, Bono, and the Gates Foundation Rock the World with OKRs\", Portfolio/Penguin, New York, 2018, ISBN 978-0-525-53622-2 (book by the author who popularized the method).
\n[3] Neely, A., Adams, C. and Kennerley, M., \"The Performance Prism: The Scorecard for Measuring and Managing Business Success\", Cranfield School of Management / Financial Times Prentice Hall, 2002. Available at: https://www.cranfield.ac.uk/som/research-centres/centre-for-business-performance
\n[4] Bititci, U. S., Garengo, P., Dörfler, V. and Nudurupati, S., \"Performance Measurement: Challenges for Tomorrow\", International Journal of Management Reviews, vol. 14, 2012. Available at: https://onlinelibrary.wiley.com/doi/10.1111/j.1468-2370.2012.00336.x
\n[5] Garengo, P., Biazzo, S. and Bititci, U. S., \"Performance Measurement Systems in SMEs: A Review for a Research Agenda\", International Journal of Management Reviews, vol. 7, 2005. Available at: https://onlinelibrary.wiley.com/doi/10.1111/j.1468-2370.2005.00105.x
\n[6] ISTAT, \"Imprese e ICT, Anno 2025\", Istituto Nazionale di Statistica (Italian National Institute of Statistics), 2025. Available at: https://www.istat.it/comunicato-stampa/imprese-e-ict-anno-2025/
\n[7] Baltrunaite, A., Formai, S., Linarello, A., Mocetti, S., \"Proprietà, governance, management e performance delle imprese: evidenze dalle imprese italiane\", Banca d'Italia, Questioni di Economia e Finanza no. 678, March 2022. Available at: https://www.bancaditalia.it/pubblicazioni/qef/2022-0678/QEF_678_22.pdf
","headings":[{"level":2,"text":"Understanding the measurement system before the individual indicators","id":"understanding-the-measurement-system-before-the-individual-indicators"},{"level":2,"text":"Deciding what to measure before choosing how to measure it","id":"deciding-what-to-measure-before-choosing-how-to-measure-it"},{"level":2,"text":"Translating goals into operational indicators with the strategic cascade","id":"translating-goals-into-operational-indicators-with-the-strategic-cascade"},{"level":2,"text":"Choosing the reference model: Balanced Scorecard, OKR or Performance Prism","id":"choosing-the-reference-model-balanced-scorecard-okr-or-performance-prism"},{"level":2,"text":"Equipping the system, from spreadsheets to integrated platforms","id":"equipping-the-system-from-spreadsheets-to-integrated-platforms"},{"level":2,"text":"Five mistakes that strip a measurement system of meaning (even a well-designed one)","id":"five-mistakes-that-strip-a-measurement-system-of-meaning-even-a-well-designed-one"},{"level":2,"text":"Limits and conditions of applicability","id":"limits-and-conditions-of-applicability"},{"level":2,"text":"FAQ — Frequently asked questions about performance measurement","id":"faq--frequently-asked-questions-about-performance-measurement"},{"level":2,"text":"Operational summary","id":"operational-summary"},{"level":2,"text":"Conclusion","id":"conclusion"},{"level":2,"text":"Sources and references","id":"sources-and-references"}],"tldr":"Should you measure everything that can be measured, or focus on a few essential indicators? There is no single answer: it depends on how the system that holds them together is built — and most companies don't ask themselves this question before choosing their KPIs.","tldrItems":null}