Purchasing & LogisticsExpert Information - 17.03.17
Spring Cleaning in the Database
How to Improve Data Quality in Seven Steps
People usually spring-clean their home once a year. Companies would do well to cleanse their data, too. After all, their success increasingly depends on the quality of their data: if they want to achieve a high level of automation, they first have to detect and eliminate incorrect and incomplete records. However, it is not sufficient to cleanse data once in a while. It has to be done on a regular basis. proALPHA explains how data quality can be improved in the long term in seven steps.
1. Assign Responsibilities
In all departments and the entire company, it has to be defined clearly which employees are responsible for entering data and which ones for cleansing them. Inconsistent data may occur, for example, if the engineering department creates purchased parts for a product in the CAD system and if the purchasing department does the same in the ERP system. Clearly defined responsibilities help avoid these pitfalls.
2. Define Quality Standards
Each company and each process place different demands on data. Is a product record only considered complete if an image has been attached? Or if the description has been translated? Such rules are crucial to analyzing data volumes. Moreover, it has to be defined how incorrect data are to be handled.
3. Analyze the Status Quo
Companies have to get an overview of the current status of their data. proALPHA offers proven analysis methods and tools to detect errors in data volumes and to determine their frequency. Incomplete or inaccurate data, redundant master files and conflicts in various data pools can be found in no time.
Many databases contain records that are either obsolete or no longer required for other reasons. Companies should therefore check whether they can archive data they no longer need in their daily business. Retention periods affecting customers and authorities have to be respected, of course.
5. Automate Data Flows
Data from various departments can be compiled or even automatically entered by means of workflows. For example, if a new customer and the corresponding master files are created in sales, a request for checking this customer's creditworthiness can be automatically sent to the accounts receivable department. At the same time, an employee in financial accounting is asked to complete the account master files. After both tasks have been finished, a notification is sent to the employee responsible for credit limits. Companies therefore do well to check whether there are steps that can be mapped with a workflow.
6. Train Employees
Technology alone is not sufficient to improve data quality. There are rules people have to follow without being supported by digital processes. This includes definitions for uniform spellings, for example, in addresses. If employees receive regular reminders and training, they will not forget about these rules in their hectic daily work. Moreover, training will sensitize employees to the importance of data quality.
7. Schedule Regular Data Quality Checks
Quality controls and data cleansing need to be done on a regular basis. Integrated data mining tools such as the proALPHA Analyzer as well as special quality control tools such as InfoZoom Data Quality provide convenient options for this purpose.
Experience has shown that simple measures may have a major effect. Costs caused by inaccurate data can be significantly reduced, for example. Companies that have full control of their data can grow faster and achieve higher sales. According to the study "Datenqualität und –management. Trends 2016 (Data Quality and Data Management. Trends 2016)", this increase in sales may amount to up to 29 percent. The rather small investment in high data quality will pay off in the blink of an eye.
Susanne Körber-Wilhelm will be happy to answer your questions and provide you with further information about proALPHA Public Relations.
+49 89 92306841-445 firstname.lastname@example.org