Data quality within company structures can be defined as data that is good enough to use for a specific purpose. The right quality of marketing data can help convincing your customers to buy a product via a certain channel. The main goal of striving to qualitative data is to stimulate data-driven decisions because errors make your information practically useless. Within this article you’ll learn more about the different aspects when talking about data quality.
The different aspects of data quality
Despite increasing importance, data maintenance and quality are often still a little bleak. It is a problem that a lot of major companies still face and that impacts their core business and operations. The last decade companies have been understanding the challenges of having/maintaining good data quality and started investing in the integration of a master data strategy into the company-wide strategy.
In fact maintaining data quality is often a complex job, because it often depends on the contexts. However, there are some aspects that can be generally used so that your company is ready to tackle most data quality issues.
The different aspects of data quality
Completeness
The completeness of data refers to the fact whether all information has been filled out. This does not mean data is validated on its correctness/accuracy, but already provides the user an overview of what is missing. Often these rules can be defined for specific data fields and it is even possible to give some of them more importance by setting a weight on the corresponding data field. Descriptions for the website might be more important than adding a color for specific product types, in this case the description gets a weight of 50% and the color only 25%. By doings so, end users are able to determine which information is more important.
Validation
Most of the PIM or MDM solutions allow users to set validation types which means only a specific type of data can be set. It is possible to set the data type of a field too eg. a list of values, which allows the users to only select information from a list. The user is not able to add any other values, so the risk of making human errors becomes more limited. It is also possible to ensure certain actions in the PIM or MDM solution cannot be performed, unless the data is up to standard of a specific validation. So, when performing an import, data is not going to be imported if the values are not conform the rules.
Uniqueness
Uniqueness validations can be setup which look to specific data types to ensure their uniqueness. GTIN’s are used for this purpose. This type of code is determined by GS1 and indicates whether your product is entirely unique based on global rules. For brand, supplier and customer information names and extra fields can be used. The user will immediately get a notification that data is not unique and that the information cannot be saved. Deduplication of data can take place in different steps of the data setup, for example in the import and in the system itself.
Consistency
Since a user is able to fill out information for different channels in one system, data is consistent over all their information sources. Users can use the same data information across all channels or use it onto the same core data set, so customers always have accurate data even if they access it via different channels.