Performance Architect update 44/2011

Advice on KPI documentation and configuration

Configuring KPIs following their selection is represented by the documentation of the complete set of relevant details for each KPI and the activation of KPIs so that data can be reported and analyzed.

  1. Link KPIs upstream with business objectives and downstream with organizational initiatives. KPIs should be connected to organizational objective as they make objectives SMART. Initiatives should be establish to support the achievement of objectives by improving KPI results.
  2. Assign a data custodian responsible for gathering measurement data for the KPI. Data gathering for each KPI requires clarity and ownership. Having a responsible for collecting KPI data is a management approach to ensure accountability with data being available for analysis on time.
  3. Assign a KPI owner responsible for the achievement of the desired results. Each KPI should have a manager allocated as its owner, to ensure responsibility regarding its analysis, results and improvement options.
  4. Avoid tunnel KPI definitions – repeating the KPI name in the definition doesn’t add value. Good practice in working with KPIs requires thorough documentation of what they reprezent. Proper KPI definitions should go beyond repeating the KPI name, by providing a plain English explanation of what the KPI is about.
  5. Categorize KPIs by their reporting status - active = data is tracked, inactive = data not available. Activating KPIs is the process of moving a KPI status from inactive, when the data is not available to active, when data is reported and a clear process is in place for doing so on a regular basis.
  6. Clearly identify the unit type, most of the time % (percentage), # (number) and $ (dollar value). KPIs being measurable entities, they have an associated unit type. To simplify communication, the symbol should be used instead of the word expressing it.
  7. Data accuracy for each KPI should be evaluated as low, medium and high and treated as such. Not all KPIs have the same data reliability. Survey based KPIs are always going to be less reliable compared to revenue KPIs, due to objectivity issues. Other aspects to be considered are data automation and auditing.
  8. Determine the frequency of data generation and the frequency of reporting for each KPI. Data for some KPIs, such as ‘# Website visits‘ can be easily gathered on a daily basis. For other KPIs, such as ‘% Employee engagement‘, data gathering requires considerable costs and efforts, impacting a large number of staff. The frequency of reporting is influence by factors such as cost, efforts and technical complexity.
  9. Develop a customized KPI documentation form that contains the relevant details describing the KPI. Documenting KPIs can be easily done in a template that structures the main description fields considered relevant for the organization. contains such a model that can be customized at organizational level.
  10. Document if the trend is good when increasing, decreasing or when data is within a range. For some KPIs the results are good when they are decreasing from a period to another - for example ‘# Customer complaints’. For others, such as ‘$ Revenues‘, the results are good when increasing, while in the case of ‘% Budget variance‘, the results are good when within a specific range.
  11. Document where the reporting data for each KPI is sourced from and who produces it. Understanding a KPI relies on having a clear understanding of the data behind it and its source.
  12. Don’t worry too much about a KPI being leading or lagging. Differentiating between the two is debatable and confusing. What is considered a leading KPI for some is a lagging KPI for others. As agreement around this differentiation is oftentimes difficult to achieve, it is secondary in importance and impact.
  13. Ensure each KPI is clearly explained in a definition and has a purpose for usage. The separation between definition and purpose is essential. The purpose expresses the reason for using the KPI and is one of the key components of the documentation form.
  14. KISS - keep it short and simple: Use the # and % symbol to replace “number” and “percentage” in KPI names. Standardizing KPI names and shortening them supports communication and enables clear data visualization of KPIs in dashboard and scorecards.
  15. Simplify KPI names by eliminating the word “of”. As a “common denominator” it can be cut from the name. KPIs are analytical in nature and where possible, their names should be as concise as possible. Definitions, calculation and purpose fields provide context and can be more wordy.

Aurel Brudan
Performance Architect,

Leave a Reply




[Name and Mail is required. Mail won't be published.]