Data Entry Deserves Your Respect
As any Lean Six Sigma White, Yellow, Green, or Black Belt will tell you, data is fundamental to identifying and improving processes. But what if much of an organization’s data is bad – or “dirty.”
Is Dirty Data Costing You Money?
IBM has estimated the costs of bad data to be $3 trillion dollars for U.S. businesses. And many executives believe that as much as a fifth of their business data is unreliable. Dirty data can be found in product records, sales and marketing data, financial information – anywhere data is collected.
Dirty data is defined as data that is incomplete, out of date, irrelevant, redundant, improperly formatted, or incorrect (including typographical errors). While the Eight Wastes of Lean were initially applied to manufacturing, a number of authors have shown how the Wastes can apply to dirty data.
- Data that is incomplete or incorrect could be associated with Waste of Defects or the Waste of Human potential. In the latter case, errors might result from untrained staff, overworked staff, or inadequate documentation.
- Data that is improperly formatted or redundant may reflect the Waste of Overprocessing. For example, the same piece of data entered into multiple systems, possibly in different ways by different teams.
- Data that is out of date or irrelevant suggests a Waste of Inventory.
Use Lean Six Sigma to Identify Dirty Data
If you are concerned about dirty data in your work, your Lean Six Sigma training can help. Identify a problem area and use the DMAIC process and LEAN tools to improve. Here are a few ideas to get started:
- Focus your efforts on a particular system. Use value process mapping to identify how, where, what, and when data is entered/updated. Also indicate who is responsible for the data entry. Look for opportunities to eliminate or minimize wasteful actions or steps.
- Schedule a GEMBA walk. When I was dealing with unhappy magazine subscribers, I asked the fulfillment service if I could drop by their office and follow an order from receipt to data entry. All seemed well until I sat down with the data entry clerk. Overwhelmed by the volume of his work load, the clerk dropped orders he had received but not entered that day into a box beneath his desk. That was his daily practice. Enter what he could, “file” the rest.
- Gather VOICE of the Customer information. Are you gathering the data they need? Are you providing too much data? Understand the value of each piece of data collected –what information it provides, who uses it, how often, and how.
- Listen to the Voice of employees and respect their input. Remember, they are your customers too.*
Use Lean Six Sigma to Eliminate Dirty Data
Using what you have learned, create and implement a plan to cleanse your data and set up processes to keep the data clean. Build in quality control and error-prevention strategies. Consider how you can apply the concepts of Jidoka (fix a problem as soon as possible) and Poka Yoke (mistake proof a process). For data, that could mean having a second person proofread records, or using software to enforce consistent formatting or recognize and reject improperly formed email addresses.
Don’t forget to document data requirements, including defintions; establish processes; train everyone, and enforce standards. Most importantly, instill a culture of respect for data and the people who enter throughout your organization.
*I was involved with a project where the data entry work of production coordinators increased by almost 100%. The extra workload lasted for over a year and the coordinators could not keep up with the many changes and updates that needed to be reported for thousands of products. Because of missing or outdated schedule data, a sister department was having trouble finding vendors to manufacture the products. And customers were not getting what they needed on time. A meeting was called to find a solution. The coordinators requested some simple changes to the system they were using and hiring temps to help catch up. Instead the VP in charge of both departments told the coordinators they needed to update a separate weekly Excel report for the sister department. This was redundant to and in addition to their regular updates. Fortunately, a few weeks later, the Excel spreadsheet was eliminated and temps were hired.