Data analytics is now a popular topic among compliance professionals. You can hardly go to a compliance conference these days without presentations devoted to the topic. As someone who has passionately advocated for the use of data in compliance, I cannot be more gratified by this development. I also recognize, however, that the talk of “data analytics” remains somewhat intimidating and mystifying for some compliance professionals.
Let’s look at some of the myths around “data analytics” in compliance.
Myth #1: Data Analytics is All About Numbers
Data is simply information, which does not exist only in numerical form. The number of investigations on your docket is a data point; the types of investigations in another data point. Chats in a chatroom are data in a textual, unstructured form; numbers in a spreadsheet are data in numerical, structured form.
Analytics is the process of examining and making sense of data. When you interpret your number and types of investigation over a certain time period to form inferences such as “our cases involving breaches of financial control are on the rise,” you are already performing data analytics.
Demystification: Data analytics is about patterns, not numbers. It is the process of examining collections of information to make sense of them, finding patterns and outliers.
Myth #2: Data has to be Big
Data analytics has often been talked about in conjunction with “Big Data,” but no rule has ever said that data has to be big. If you are a large global corporation with tens of thousands of employees and vendors, you have big data; if you are a small business with 30 employees, you have small data. Each organization makes use of its data on a scale that makes sense for itself. A large company that processes thousands of sales commission payments per month may need software programs to analyze these payments, automatically detecting patterns and flagging outliers. A small company that makes five commission payments per month can rely on one or two people in accounts payable to do the same. The volume of payments (a data point in itself) dictates how those payments (data) should be analyzed (analytics).
Demystification: Data is simply information. There is no size requirement for analysis.
Myth #3: Data Analytics Requires Artificial Intelligence (“AI”)
Similar to the association with “Big Data,” data analytics is often lumped together with AI and machine learning. Yes, when you have truly big data, you can do some very fancy things with it by leveraging AI and machine learning capacities, but the vast majority of companies do not need that level of sophistication to improve performance. You do not need artificial intelligence to examine data points such as revenue, employee populations, the number and categories of third parties and payments, or the volume and types of business activities and transactions in order to develop the risk profile for your company: You just need accurate information and common sense in putting it all together.
Demystification: Data analytics requires accurate information and common sense, not artificial intelligence.
Myth #4: You Have to be Good at Math/Statistics to Understand Data Analytics
I am living proof that you can be bad at math and statistics and still embrace data analytics. I avoided all math classes and nearly flunked my statistic course in college; yet I appreciate data analytics because they provide precision and indication. When I worked with corporate monitors in the Fraud Section at the U.S. Department of Justice, I was known to say: “Use data instead of adjectives.” When someone says: “Credit card transactions in this subsidiary are rare,” I don’t understand what that means exactly; but if she says “12 percent of all employee expenses in this subsidiary are through credit cards,” I would understand precisely what that means. Posters of a hotline may be pretty, but it is data such as the number and types of reports received that actually indicate hotline usage.
Demystification: You do not need to play a musical instrument to appreciate music, and you do not need to be a mathematician or statistician to appreciate data analytics.
Myth #5: Data Analytics is Expensive
Actually, not analyzing your data can be expensive. I have seen companies building third-party programs disproportionate to their third-party risks, requiring training for all employees for issues touching only 10 percent of them and adding controls that impact thousands of transactions when only dozens of them are at risk. Many of these burdensome and costly investments might have been avoided if relevant data had been analyzed and considered. Business leaders have been using data analytics to reduce cost and improve productivity for decades, and they would expect their compliance counterparts to do the same.
Demystification: Data analytics requires only that you understand what information you have and what sense careful interpretation can make of it for your organization’s specific purposes. Committing to data collection and interpretation is not nearly as expensive as not doing it.
Hui Chen is an independent ethics and compliance consultant and was the Justice Department’s first-ever compliance counsel expert. She had served in global senior compliance lead positions at Microsoft, Pfizer, and Standard Chartered Bank.