As students return to school, there are some predictable changes in internet traffic as well as some clear differences with last year. I looked at category traffic levels for all of the parent categories that we track at Hitwise along with some of the larger sub-categories (Search Engines, Social Networking and Forums and Email). The analysis provides insight into online seasonality as well as some changes as a result of the election, hurricanes and the economy.
Changes in Behavior Year on Year

At this time last year we saw an increase in visits to Automotive websites whereas they are flat this year. Visits to Shopping and Classifieds websites are flat this year while they increased 8% last year and visits to Travel websites have fallen more than normal, down 15% this year compared to a 7% decline last year. All of these changes reflect the struggling economy.
Seasonal Patterns Year to Year
The most obvious seasonal pattern is a decrease in visits to social networks and a corresponding increase in visits to Education websites. Comparing the first week of August to the first week of September, US visits to Social Networking and Forums websites are down 8% while visits to Education websites are up 16% . We saw a similar pattern at this time last year. This is a direct results of kids going back to school.
Take a look at this chart – it shows visits to MySpace on the left axis and visits to websites in the Education category on the right axis. A beautiful illustration of seasonal off-sets if I’ve ever seen one!

Visits to News and Media are up 17% in the past month. I thought the lift was a result of the hurricanes, the RNC and the DNC. But these events are only exaggerating a natural seasonal lift. Visits to News and Media websites increased 12% in the same period last year. Perhaps as we all buckle down to work, we pay more attention to the news.
This is a great reminder to marketers of the need for external data sources to put any decline in context as well as allowing for year to year comparisons across a sector.