Data Collection Plan

Skill level: Basic to advanced


A critical activity for process improvement, data collection can be performed on a continuous basis or for a short period of time. There are many different ways of collecting data, and each is designed to serve a specific purpose. The data collected can be analyzed using a very large number of tools, from Pareto charts to hypothesis testing.


  • Provides for quantification of process performance from which data-driven decisions can be made
  • Provides structure and consistency around data to be collected
  • Data can be variable (continuous) or discrete (attribute)
  • Provides a common language that enables learning

How to Use

  • Step 1.  Develop a data collection plan based on the process map and priority matrix.
  • Step 2.  Develop the data collection tool and test it. Some typical components are:
    • Name of measure (speed, cycle time, accuracy, etc)
    • Type of measure (input, process, output)
    • Type of data (variable or attribute)
    • Operational definition (enables common understanding)
    • Specification (least acceptable performance)
    • Target (ideal performance)
    • Type of form (needed to collect data, such as check sheet)
    • Sampling requirements (what level, if any)
  • Step 3.  Review data and correct data collection sheet or tool as needed.
  • Step 4.  Compile data in a worksheet or any statistical application available.

Relevant Definitions

Sample size: The number of samples that will be taken at a given time.

Frequency: The regularity of when data is collected (e.g., once a day, every hour).

Factors: The inputs or outputs of interest (e.g., time of day, department, time to process).


A manager of an accounting department seeks to track how long it takes to process and approve expenses reports on a weekly basis. Because the reports are processed in batch by region and not individually, each person in the department tracks the number of hours per day and the number of reports processed during that time.

Key inputs:

  • Date
  • Region from which the reports come
  • Number of reports processed
  • Total time in minutes spent to review and approve the reports



« Back to Glossary Index