Back to Basics: Forecasting the Revenues of Internet Firms

This paper examines the roles played by past revenues, web usage data, and analysts in forecasting the future revenues of internet firms during the years 1998 to 2000. For this time period our analysis shows that estimates of web traffic growth have significant incremental value in the prediction of revenues above time-series forecasts. Furthermore, analysts almost always underestimate the revenues of internet firms. Historical revenue growth has incremental predictive power over analysts' forecasts for portal and content/community firms, but not for our e-tailer sample. Moreover, the stocks of the portal and content/community firms with high historical revenue growth earn higher abnormal returns during our sample period than do those with low historical growth. Estimates of web usage growth generally do not have incremental value over analysts' forecasts for predicting the revenues of either set of firms. However, perfect foreknowledge of actual web usage growth would provide incremental predictive power. Collectively, our findings point to the potential value for forecasting purposes of both improving upon the web usage estimates and obtaining more timely reports of actual web traffic.