Today, more than ever before there is talk about Big Data and Analytics. Information sources abound and that is why Big Data can seem both needed and overwhelming. When the data seems mind boggling then the application of Analytics to understand and interpret it is impossible to fathom. Data Collection Planning is a systematic approach to focus getting what you want from the data by capturing what is needed to answer the questions that are paramount for your organization.
There are two purposes for collecting data and the application of analytics to that data. First, develop an understanding of where you are now. Second, determine the consequences of your actions on your organization.
To gain the most from Analytics requires the collection of the right data. If we view the output of the organization to its customers as a function of the myriad of inputs along the many processes that comprise the supply chain it is easy to view this as Big Data. Sources of data include:
- The organization’s outputs from the customer’s viewpoint.
- Policies, procedures, and work instructions that govern supply chain processes.
- The tools and materials that are applied and consumed.
- An organization’s strategy and its communications.
- The measurement system that is employed to capture and collect this vital information.
Achieving consistency in the collection and evaluation of the data requires the use of Operational Definitions. An effective operational definition describes in detail the characteristic that is measured, how the measurement is done, and how to make a decision either good or bad about the measured characteristic. The following acronym simplifies what makes or breaks the Operational Definition, which is U-SMART.
- Useful to you and your customers.
- Not Ambiguous
- Repeatable and Reproducible
- Terse and to the point making it understandable by all.
The best way to begin your data collection planning is to think through a series of questions. The following should provoke some thought.
- What is the purpose for collecting the data?
- Where will the data be collected?
- What kind of data is needed?
- How much data do we need to make decisions?
- How often will the data be collected?
- Who will collect it?
- What are the measurement tools and methods?
- How reliable are the methods of measurement?
If there is no stated purpose for collecting the data you are not going to find the purpose within the data. It just doesn’t work that way. A purpose and a plan are worth the effort to get the biggest bang for the analytic buck.
The following eight steps are your guide to Data Collection Planning for successful Analytics.
- Develop a list of questions about your customers, processes, and organization that you want answers to. These questions define the purpose for collection and analysis of the data.
- These questions, to be answered in a data driven way, aim you at the type of data required to get those answers.
- Now you have to decide how you will measure the data that leads to the answers of the questions.
- Determine the stratification factors and identify any environmental concerns. Stratification factor examples are location, department, city, state, country, demographics, day of week, month of year, season, and more. From an environmental standpoint think about temperature, humidity, and pollution to name a few. These variables may influence how your organization performs, but if you don’t capture them you can’t analyze them.
- Developing a sampling plan answers the question about how much data you will need.
- Decide where and how the data will be collected.
- Complete a Measurement System Analysis and define the controls that will guarantee data integrity. If you data aren’t reliable your analytics won’t be either.
- Last, but not least, is the logistics plan. Specifically outline the Who, What, Where, When, and How the data will be collected.