The primary limitation to Benford’s Law is business data do not always follow
natural patterns; there exist a large number of reasons that transactions may not match
Benford’s Law. Explanations like recurring fixed expenses, unusual business cycles, and
assigned amounts are often found. The author has taught digital analysis to thousands of
professional auditors; in ten years of asking participants about their success with digital
analysis, only three individuals have reported finding fraud with Benford’s Law (others
have reported that digital analysis could have been used to find already discovered frauds,
but hind sight is not prediction). In some ways, the audit field may have overestimated
the usefulness of digital analysis. But despite its limitations, Benford’s Law remains one
of the most popular data mining techniques for fraud.