The State Of Data Quality In 2020 – Everything You Must Know
Melissa IN Team | |
Today, data is an organization’s most valuable asset. The way this data is managed and analyzed is changing the way business information is used. This has made people at all levels of enterprises alerted to the importance of data quality. Unfortunately, when a survey was conducted, very few organizations surveyed had dedicated data quality teams. The organizations were found to be dealing with multiple data quality issues and lacked the building blocks of data management and governance. This could have a negative impact on budget management, communication with consumers, marketing, sales, etc.
Common Data Quality Issues
There are several areas that contribute to poor data quality. Key amongst them are:
Absence of Standardization Norms
This is one of the most common data quality issues faced by organizations. There is no lack of raw data available. However, this data comes in from multiple sources. The most common channels for data collection include sales representatives, websites, mobile apps and call centers. In many cases, it does not usually follow a standardized format. It can thus give rise to inconsistencies. For example, at one point of data entry, a date may be entered in the DD/MM/YY format while at another it may be entered in the MM/DD/YY format. Similarly, a customer’s name may be entered with different spellings. This can not only cause confusion but lead to duplication of records.
Poor data quality control at entry level
Many organizations still rely on manual data entry. This has a high risk of error and is one of the leading causes of poor data quality. The person may enter a wrong value, create duplicate records or may even miss a field. Without proper quality control at this level, the data will go forward and affect all other records as well.
For example, when creating a daily sales report, the person may enter the number of products sold as 20 instead of 200. If not checked, this can affect reordering supplies, the company’s calculation of profits, etc. Poor quality control at data entry levels is responsible for common data errors such as incomplete data, outdated information and inaccuracies.
Poor quality data from third party sources
Not all data used by businesses is generated in house. A lot of data is taken from third parties. This data is often subject to inaccuracies, outdated information, incomplete data and other such common errors. For example, companies may look at third party survey results when planning a marketing campaign. This limits their control over data quality as they may get only a partial view of the survey. Many companies are beginning to realize this and are turning towards government agencies for reliable third party data. This is important especially when data is needed to verify identities.
While there is no dearth of data available, not all of this data is structured into usable form. Without proper labels and tags, it is very difficult to use this data. For example, you may have a phone number and address but if this is not connected to a name, it is quite useless. There is also a lack of metadata such as date the data was created or modified, the author, etc. Without this metadata, it is hard to assess how reliable the data is.
Lack of Internal Communication
Most businesses use 3-4 channels to collect and compile data. The data collected by one team is used not only by them but by others as well. For example, data about the density of consumers in different parts of the city may be used by the marketing team as well as the delivery team. However, these teams often do not communicate with each other. For example, the delivery team may realize that phone number associated with a particular address is wrong. They may correct the number in their own records but may not share this with the other teams. Thus, if a third team wants data about the customer, he will get two records- one with an old, invalid phone number and one with the current number.
Absence of a Centralized Data Quality Strategy
A majority of organizations still do not follow a centralized data quality strategy or a team responsible for data quality. As a result, there is no one to take ownership of data. Individual teams are left free to take their own decisions regarding how to collect data, how to tabulate it and what quality standards to maintain. But, what measures as good for one team may not be good enough for another. For example, the pin code may not be a very important data field for the final delivery team but this could be crucial for the marketing team.
There are a number of other issues that plague data quality. Many organizations simply do not have the relevant technology for data quality validation. In addition, there may be a lack of internal manual resources and an insufficient budget to improve upon them. Another common issue is inadequate support from senior management. Though they may use data, they do not take ownership for it and hence do not see the point for enhancing it.
Data quality is an issue that cannot be ignored. The challenges are numerous but they can be overcome. All organizations, big and small need to address issues related to data quality and take formal steps to improve upon them. Creating a dedicated data quality team is crucial to this. Artificial Intelligence initiatives may also be used to identify data quality issues and enhance the data quality. AI-enriched tools can also improve data efficacy and increase productivity. When it comes to data quality, it is important to note that data enhancement is not a one-time effort but needs to be an ongoing process.