By now, everyone knows that organisations can own an awful lot of valuable data and gain many useful insights from it. For this reason, many organisations choose to launch various data initiatives, ranging from monitor dashboards to predictive data models. However, good data quality is essential to get the insights from data right. Those working with the wrong data will also see wrong outcomes. All the more important to keep a finger on the pulse on data quality. In this article, we cover five key signs that your organisation may be struggling with (a lack of) data quality.
Signal 1: Inconsistent reporting
Reports are often used to provide insight into developments of, say, an entire company, a department or a project. When the report contains inconsistencies, one does not know which information is correct. This can make it difficult to use the data for (re)guidance. Examples of reporting inconsistencies include data in different reports not matching each other or monthly reports differing from the annual summary. Another common example is KPIs reported in different periods that show significant fluctuations for no apparent reason.
How to deal with inconsistent reporting? By setting automated controls and systematically monitoring patterns, your organisations can spot inconsistencies early. Underlying data quality issues are thus directly addressed.
Signal 2: Discussions on definitions
When the same data are captured in multiple ways (e.g. in different systems), discussions often arise about possible differences in definitions and which definition should be considered leading. Interpretation problems arise when data are not captured according to fixed standardised norms. Discussion then arises about the correct definition, the use of the definition and the meaning of the data.
The lack of a standard can also lead to data entry errors, as employees may use different notations or not understand exactly what data should be entered in specific fields. A lack of uniformity complicates data comparison and analysis. This leads to confusion. Besides causing confusion, a lack of uniformity can limit and delay the ability to extract meaningful insights from data.
Solution: Implement clear data definitions and standards, clear data governance and matching protocols. Linsey and Jelmer are happy to share their experience on setting up data governance processes.
Signal 3: Delay in processing data
Slow processing of data can hinder an organisation's ability to respond in a timely manner. Examples where this occurs include situations where real-time processing is essential, such as in financial transactions or monitoring of operational processes. Delays can then have serious consequences. Also, when data insights are requested and it takes a long time before they can be provided is something that can be linked to slow data processing. People know the data is available, but it takes a long time before it is made available.
Data availability is an important part of data quality. Good data quality implies that the required data is available when it is needed and that it is accessible to those who use it without unnecessary delays or obstacles. An efficient business process requires fast and reliable data processing. This signal indicates potential bottlenecks in the data infrastructure, such as outdated systems or insufficient capacity.
How to speed up data processing? To address this signal, it is important to invest in a data infrastructure that enables fast (real-time) data processing.
Signal 4: Excel as the go-to for storing important data
Many an organisation is familiar with it: the large Excel files packed with important data of which every colleague seems to have a different version. Everyone knows the file, but no one is familiar with the administrator.
Storing important data in Excel files (shadow IT) often indicates a lack of clear rules, procedures and/or responsibilities for managing and safeguarding the data. As this data is not standardised and controlled, the likelihood of incorrect data increases.
How do you solve this? Reducing Shadow-IT and integrating this data into formal systems are important steps an organisation can take to improve data quality.
Signal 5: Regular need for manual data cleansing
When urgent data cleansing is called for with great frequency, this is a signal that A) these are important data objects and B) there are fundamental data issues for the objects. The frequent need to correct data indicates an underlying problem in data quality.
Spend less time on manual cleaning? Identify the causes of these errors and implement automated processes to reduce the need for manual corrections.
Do you recognise one or more signs within your organisation? Improven is ready to provide further support and guide your organisation on the path to Data Excellence.
We have a strong track record in helping organisations, including through:
- Performing the Improven Data quality quick scan
- Creating a roadmap for improvements
- Establishing and securing KPIs with regard to data quality
- Optimising and securing data quality processes.
Want to know more about Improven's Data Excellence? Above all, read continue.
For more information on this article, contact Hugo ter Welle on 06 45 18 47 09 or hugo.ter.welle@improven.nl