Data quality in Human Ressources services
Companies’ HR services are facing data quality challenges more and more frequently; Unfortunately, they often notice it too late.
“In IT, data quality refers to data’s conformity to predicted usage data in modus operandi, processes, decision-making and planning.”(J.M. Juran).
Furthermore, data quality can be evaluated according to the more or less accurate representations of the production process they come from.
Data quality is based on 5 basic principles:
When the data do not follow these five principles, HR services have to face numerous difficulties resulting from bad data quality in their system.
The best practices to reinforce the information systems are evaluated in terms of quality improvement, time saving and reduction of anomalies management costs.
How can these theoretical attributes slowdown your information systems efficiency, cause blockings and compromise your credibility?
Let’s now look at some common examples, related to bad data quality.
Nowadays, HR services have to treat a lot of different data. However, this abundance can sometimes be problematic, because it can be very complex to make sure that all required data are present in the system, just like the data treatment process. Respecting the completeness principle allows to make sure that no information is left out of the system, or that no present data is incorrect. Not respecting the completeness principle will result, for example, in a lack of data regarding the attendance hours of some employees, essential to calculate their salary, which can result in undue payments. Another example is the lack of information regarding drivers’ driving license validity. This could also engage the company legally as a serious fault due to gross negligence. Therefore, it is very obvious that having incorrect or incomplete information during the monthly data treatment can have more or less significant consequences.
According to the company’s industry, business, size and its environment, we can define from experience data standards, allowing us to get a general idea of the conformity of the results. In that case, we talk about the plausibility of results. Plausibility refers to what is usually admitted for a data, according to its consistency with business data. To guarantee data’s plausibility, we verify compliance between treated data and business data standards. This verification is performed in order to avoid discrepancies and so a loss credibility of HR services regarding their data standards knowledge in their field. For example, the HR service should be able to verify the likelihood of classic seasonality for days off taken in the company, compared to vacation months.
A company has to manage a large number of data, often linked to some unsynchronized referential systems.
But how do we make sure that the key data collected in third part systems are correctly referenced in our system?
It is then about testing the referential integrity. Integrity allows us to make sure that data quality is not corrupted because of a non-correspondence in the referential system. It is then important to check the regularity and the correspondences completeness with these referential systems, and to make sure that they are well synchronized.
An integrity problem may especially occur when the “Employees” referential system is not synchronized with the HR system and the leave tracking system. For example, the leave referential system could use the identification “Employee” that does not exist in HR system. Likewise, some duplicates of the same element may be found. We notice here some data integrity problems, but also a waste of time for the HR service trying to solve these discrepancies.
An automatic control of the referential system allows to identify the issues beforehand and to ask the business managers the required updates; or even to process automatic updates, if necessary.
HR services often have to manipulate a multitude of data from different services and produced by different departments. Here the difficulty is to control data consistency and the absence of any contradictory data. We speak then about data consistency. Consistency refers to the lack of contradiction between a same data presented by two different systems. For example, consistency problems may appear comparing information from the HR system with the “Group Reporting” or the “Financial Reporting” (FTE, worked days, payroll…). We can easily imagine the impact that contradictory results could produce on an HR service. If HR service lets the inconsistent data pass through a report, it would be very complex to distinguish which data is accurate for use.
It has become more and more complex for HR services to calculate indicators. These are difficulties that currently occur when trying to establish clear salary slips understandable by everyone, while being able to explain each calculation in detail. Simultaneously, various regulations highlight the lack of transparency regarding employees. This opacity is the cause of various suspicions concerning HR services reliability. Reliability is the ability to logically demonstrate the current data, because the calculation methods used to produce this data have been evaluated and verified, in order to generate data faithfully conformed to reality. Concretely, it’s about verifying the validity of business management rules to prove that the measures work well. To calculate bonuses or overtime in order to ensure that paid amounts by company and amounts owed are consistent, guarantees reliability. Obviously, a company which is paying undue payments, or conversely, forgetting some obligations, and is not able to verify its calculation, compromises its reliability and credibility. Consequently, the validity of the systems used to produce data is questioned. That is why, both from a business and regulations point of view, we are going towards systems automation in order to avoid reliability issues.
Fig.5. Comparison between granted bonuses and calculated bonues thanks to Quality-Gates
To conclude, the amount of data that HR services have to manage every day forces them to implement important quality control processes.
These processes are necessary for the company in order to avoid anomalies, but they are often complicated and expensive.
The Quality Gates tool answers the challenges created by quality management, especially in HR services by matching the requirements of the business. It allows automating a number of controls and alerts.
Quality Gates offers HR managers the opportunity to simply comply with the main requirements they are facing: completeness, plausibility, integrity, consistency and reliability.