Audit Analytics - What is it really and who makes a good data analyst?

Audit Analytics

What is it really and who makes a good data analyst?

Over the course of my career, I have had the opportunity to participate and lead teams where there was a heavy need for analytics across many different client engagements and in many different environments.

At the beginning of my career Computer Assisted Audit Techniques (CAATs) was the term used to describe procedures used to verify the accuracy of certain calculations and reports used by clients that including AR aging and journal entry testing. The tools used then were fairly rudimentary and were constrained by available computing resources. These procedures were primarily used in fairly basic validation procedures.

Fast forward to 2015, and data analytics  or data science are the en vogue terms that connote a broader field than CAATs or other audit analytics.  This has led to a wide variety of interpretations of what analytics are and has been particularly acute for me as I interview candidates for a data analytics position open on the Data Analytics team I lead.

One recent definition of data analytics stated the following:

Analytics is the discovery and communication of meaningful patterns in data. Especially valuable in areas rich with recorded information, analytics relies on the simultaneous application of statistics, computer programming and operations research to quantify performance. Analytics often favors data visualization to communicate insight  (emphasis added).  I would substitute "operations research" for "a certain amount of domain knowledge."

I find that most candidates, whether those professing data analytics experience or those newly minted graduates of data analytics programs, lack sufficient knowledge in statistics, computer programming, databases, and/or data visualization.  I have worked with several very capable and intelligent people; however, they tend to come from computer programming, and to a lesser extent, database administration backgrounds.  These are important and highly desirable skillsets in information technology, but often are not sufficient.

Implicit in the definition is the need to actually do analysis.  Frequently, the idea of performing analytics seems to be producing data in a slightly different form than what is already available.  I see very limited analysis of data to identify trends in data. I believe this is driven by two causes--(1) a lack of understanding by the data analyst of what the data means and (2) the inability to effectively communicate what is discovered.  In addition, in order to have effective analytics in an Internal Audit department, one must have an intellectual curiosity about the area from which the data is pulled. I believe this is where many miss opportunities to add value.

All too often, the analytics teams are put into a silo where the audit teams give them a list of things to produce or perhaps are given a set of analytics "tests."  Rarely, have I seen a team that is truly integrated with members from financial and IT audit and data analytics working together starting with the planning phase of a project to can brainstorm what are the items with the greatest impact to the project and, potentially, of greatest risk / interest in the eyes of management.