Improve your audit performance quick & cost-effective
Full population testing instead of audit sampling
Author: Katharina Laumann, Content Manager and Editor at PAF
In any company auditing is important because it ensures that the company’s processes, requirements, policies or financial records are correct and meet the required standards.
Internal audits are conducted by a specially trained auditor, often as part of a Quality Management program.
A distinction is made between two types of audits:
- In the area of static quality management, audits serve as proof of contractual agreements. They are therefore carried out only once per review cycle.
- In dynamic quality assurance, audits serve to identify trends and provide the initiators of change with important feedback on the effectiveness of the measures introduced.
Get all the latest news about PAFnow and Power BI, find out what events and webinars are coming up, get helpful videos, etc. >> Register here
Audits are also frequently used to identify general problems or improvement potential and to monitor the effectiveness of improvement measures.
Regardless of the type of audit, auditors always face the same challenge: due to time and money constraints, it is nearly impossible to review every record, so auditors must rely on sampling in the most critical areas.
The Risks of Audit Sampling
- The tested attribute or assertion is judged to be unlikely, when in fact it is likely.
- The attribute or assertion tested is judged to be likely, when in fact it is not.
In addition, there is the possibility of an error that occurs because a single isolated event occurs in the sample that is not representative of errors in the population.
The most effective way for auditors to reduce sampling risk is to increase sample size.
But how can you test a larger sample or even better the whole population without making the audit process uneconomical and impractical?
The answer to that problem is: Full Population testing with Process Mining!!
How does it work?
Process Mining tools use the data that is stored in a company system to create a visual representation of any business process. The clever thing about Process Mining is, that it groups all cases that follow the exact same path together. So all the routes that a process has taken in a company become immediately visible.
This allows auditors to find outliers at a glance. But Process Mining doesn’t stop there.
The data can be broken down further, for example based on lead times, rework or transaction volume. If deviations occur, AI-supported root cause analysis finds and lists the most likely causes of that deviation.
There are also features that allow to check for compliance and violations such as segregation of duty.
Auditors can also perform a conformance check either by loading a process model into the Process Mining tool or building the model inside the tool and compare it with the as-is process.
Finally, there are also benchmarking functionalities to compare different processes and their effectiveness based on factors such as year, country, vendor, etc.
Auditors who use Process Mining drastically reduce the time needed to look at huge amounts of data while increasing the hit rate of their findings. On top, Process Mining already provides them with strong visualizations to share and prove their findings, and with clear indicators of why deviations occurred and where to look to improve processes.
Since it is so fast to perform even a full population test with Process Mining, auditors can also use it in dynamic quality assurance to monitor and benchmark improvement measures and impacts.