User Behavior Can Be Inferred From The Numbers
It is interesting to track the website traffic to one of my company’s websites. The domain, turnpikeinfo.com, is a travel site that provides a host of information about the toll roads in the United States. On any given day, by reviewing the hourly traffic, you can see when the morning rush hour begins, when the lunch hour is going or when the afternoon race home gets underway. But the Thanksgiving traffic on TurnpikeInfo.com has always been fascinating to watch. What is particularly interesting about the hour-by-hour breakdown for the turkey day event is one’s ability to draw conclusions about when people were actually eating their Thanksgiving dinners.
Our holiday traffic is always much lower than the traffic we regularly see during weekdays. It resembles the weekend traffic numbers during non-summer travel seasons; however, the figures we watch are very revealing. As the chart below may demonstrate, travelers on the road during this year’s holiday were most likely with family and friends between the hours of 2 p.m. and 7 p.m. As people headed home for the evening, the traffic spiked tremendously, especially by 9 p.m. and 10 p.m.
Looking at numbers from the 2012 Thanksgiving holiday, the hour-by-hour traffic breakdown is equally revealing. One could conclude, from the numbers, more travelers were at their destinations between about 11:00 a.m. and 6 p.m. during the 2012 Thanksgiving holiday. From that year-over-year comparison, the numbers would seem to suggest the holiday gatherings were happening a little earlier in the day during 2012.
While the charts are interesting, of course, they are hardly conclusive. But it is quite fascinating to examine visitor numbers for a website and draw educated inferences about user behavior off the site.