Real-time revenue and firm disclosure

Practical issues to consider when working with big data

“…The largest contribution of using big data comes when researchers can address questions that were previously unanswerable with more traditional data sources. As a case in point, Blankespoor et al (2022) in this issue uses real-time transaction data to estimate the performance information available to managers within a given fiscal quarter. They aggregate individual credit and debit card transaction to the firm-week level so that they can estimate a firm's performance during the quarter as of a given week.…”

“…Researchers often end up with a surprisingly small number of subjects when the data is aggregated to the unit of interest (e.g., to the firm level). For example, Blankespoor et al (2022) begins with 1.6 billion individual transactions but aggregates these to measure the performance of only 243 firms, all in the retail sector. The result is that studies using big data can suffer from a lack of statistical power.…”

“…Studies that wish to answer research questions about broad economic phenomena need to ensure that the economic forces of interest are not affected by unique attributes of the sample in which these research questions are tested. Blankespoor et al (2022) uses a sample of retail firms to document that managers respond to real-time information by suppressing negative information at the beginning, but not the end, of the fiscal quarter. However, if disclosure responses to real-time information vary across industries, then the results of this study may not generalize.…”

“…In addition, research is increasingly making use of social media data: company and customer tweets, friend networks, and content likes, to name a few (e.g., Lee et al 2015). Location and movement data captured by cell phone signals, satellite images, and taxi data can approximate physical movements of customers, shipments, or investors (e.g., Kang et al 2021) while transaction and point-of-sale data can approximate what is being purchased, how much, and when (e.g., Blankespoor et al 2022). Similarly, digital footprint data from web search and browsing, app usage, and online transactions can track users' online movement and behavior (e.g., Froot et al 2017).…”

“…In Sect. 2, I outline the four key issues to consider when working with big data and apply this approach to Blankespoor et al (2022).…”