On the Use of Consumer Tweets to Assess the Risk of Misstated Revenue in Consumer-Facing Industries: Evidence from Analytical Procedures

Abstract

We examine whether consumer-generated tweets about purchases (interest) and sentiment are useful in assessing the risk of misstated revenue in the planning stage of the audit, as reflected in improvements to analytical procedures, for firms in consumer-facing industries. We obtain consumer-generated tweeting activities from 2012 to 2017 for 76 companies in 20 consumer-facing industries from a data provider. We find that relative to a benchmark model, Twitter consumer interest, but not consumer sentiment, improves the prediction and error detection ability of analytical procedures for most firms in consumer-facing industries. Our findings are robust to different model settings. In additional tests, we observe that the effect of Twitter consumer interest is more pronounced in smaller industries and that it remains useful in analytical procedures when compared to firms’ advertising and employee headcount. Together, our results suggest that this new source of information improves auditors’ assessments of the risk of misstated revenue.