When designing measurement systems, methods, and metrics, you should always think in three layers: why you measure, what you measure, and how you measure.

Core things you must consider

  1. Purpose and alignment
    • Clear objectives: What decisions will this measurement support (e.g., quality, productivity, customer satisfaction)?
 * Strategic alignment: Metrics must reflect business goals, not just what is easy to count.

 * Stakeholders: Who will use the numbers and what behaviors do you want to encourage?
  1. Measurement quality (the “science” part)
    Your system, methods, and metrics must be technically sound.
 * Accuracy and precision: Results should be close to the true value and consistent when repeated.

 * Resolution: The measuring device must detect small enough changes to be useful (often at least 1/10 of the process variation or tolerance).

 * Repeatability and reproducibility: Different operators or repeated trials should give similar results.

 * Linearity, stability, and bias: Performance of the system should be consistent across the whole range and over time.

 * Measurement uncertainty: You should understand and, where possible, quantify how much error may be present.
  1. System design and structure
    • What you measure and where: Metrics must match the level of aggregation (team, process, organization) and the context.
 * Standards and targets: Clear reference values or tolerance limits to judge performance.

 * Frequency and timing: How often and when you measure, so you get timely information without creating unnecessary burden.

 * Reporting formats: Information must be understandable, relevant, and easy to interpret for decision‑makers.
  1. Methods, procedures, and protocols
    • Defined methods: Exactly how to measure must be documented so different people do it the same way.
 * Clear definitions: Agree what each metric means (e.g., what counts as a “defect” or a “graphic”) to avoid inconsistent counting.

 * Standard operating procedures: Step‑by‑step instructions and training for appraisers to ensure consistent execution.

 * Sampling strategy: How you select samples (random, representative) so the metric is not biased by “convenient” data.
  1. Context, environment, and constraints
    • External factors: Market conditions, competition, number of product‑markets, and regulatory requirements can shape what and how you measure.
 * Internal factors: Organization size, structure, existing tools, and practices influence the feasible design.

 * Environment and operator effects: Temperature, noise, human handling, and other conditions that can affect readings.

 * Cost and effort: The system must deliver value compared with the cost and overhead of collecting and processing data.
  1. Behavioral and ethical effects of metrics
    • Gaming and distortion: Poorly designed metrics can encourage people to optimize the number instead of the underlying system.
 * Fairness and non‑corruptibility: Metrics should not systematically favor or disadvantage groups and should be hard to manipulate.

 * Coherence: Different metrics should not conflict or pull behavior in opposite directions.

 * Diversification: Using a balanced set of indicators (not just one “golden” metric) reduces the risk of perverse incentives.
  1. Practical usability
    • Simplicity vs. detail: Measures should be as simple as possible while still capturing what matters.
 * Availability and immediacy of data: Can you access data reliably and quickly enough to support action?

 * Tool support: Availability of systems and software to automate collection and analysis where appropriate.

 * Maintainability: The system should be sustainable as processes, products, or strategies change over time.
  1. Validation and continuous improvement
    • Initial validation: Test that the system, methods, and metrics actually reflect reality and support the intended decisions.
 * Monitoring performance: Regularly review whether metrics remain relevant, stable, and useful as conditions change.

 * Feedback loops: Use experience and stakeholder feedback to refine definitions, thresholds, and procedures.

Quick HTML table summary

html

<table>
  <thead>
    <tr>
      <th>Area</th>
      <th>What to consider</th>
    </tr>
  </thead>
  <tbody>
    <tr>
      <td>Purpose & alignment</td>
      <td>Clear objectives, link to strategy, stakeholder needs, desired behaviors.[web:1][web:4][web:7][web:9]</td>
    </tr>
    <tr>
      <td>Measurement quality</td>
      <td>Accuracy, precision, resolution, repeatability, reproducibility, stability, uncertainty.[web:1][web:5]</td>
    </tr>
    <tr>
      <td>System design</td>
      <td>Choice of metrics, aggregation level, standards/targets, timing, reporting format.[web:1][web:7][web:9]</td>
    </tr>
    <tr>
      <td>Methods & procedures</td>
      <td>Documented protocols, clear definitions, sampling rules, appraiser training.[web:2][web:5]</td>
    </tr>
    <tr>
      <td>Context & constraints</td>
      <td>Market and regulatory context, organization size/structure, environment, cost and effort.[web:1][web:7][web:9]</td>
    </tr>
    <tr>
      <td>Behavioral effects</td>
      <td>Risk of gaming, fairness, coherence across metrics, need for multiple indicators.[web:4][web:7]</td>
    </tr>
    <tr>
      <td>Practical usability</td>
      <td>Simplicity, data availability, tool support, maintainability as things change.[web:3][web:4][web:7][web:9]</td>
    </tr>
    <tr>
      <td>Validation & improvement</td>
      <td>Initial validation, ongoing monitoring, feedback to refine system.[web:1][web:4][web:5][web:7]</td>
    </tr>
  </tbody>
</table>

A short illustrative example

Imagine you are designing a customer support measurement system.
You might define metrics like “first response time,” “resolution time,” and “customer satisfaction score,” ensuring the definitions and sampling rules are explicit so everyone measures them the same way.

You then check that the tools can capture timestamps accurately, that data is reported weekly in a clear dashboard, and that targets support your strategic goal (e.g., fast but also high‑quality responses) without encouraging support agents to close tickets prematurely just to look good on the metric.

TL;DR:
When you see a question like “which of the following should be considered when designing measurement systems, methods and metrics?” the best answer options are the ones that mention: alignment with objectives, technical quality (accuracy, precision, repeatability, uncertainty), context and cost, clear procedures and definitions, behavioral impacts (fairness, gaming), and ongoing validation and improvement.

Information gathered from public forums or data available on the internet and portrayed here.