Four wonders of the web

The art of web analytics has solved many of the mysteries behind customer behaviour, but some wonders of the web remain. This year, I aim to unravel the secrets behind four of them and you are invited help with the research.

Together, we can crack the mystery of cookie deletion. If users are routinely cleaning out cookies, you could find that you have significantly fewer unique visitors than you previously thought. Indeed, a study by Jupiter Research in 2005 concluded that 10% of users delete their cookies daily, 12% do so monthly and a whopping 17% remove cookies every month. Our own studies with WebAbacus have compared the number of randomly generated permanent first-party cookies with the number of registration IDs, which is a good way to tie cookies to real people. The ratio over a month has been about 1.179 cookies per ID, but this has been increasing year on year, suggesting that more people are deleting their cookies on a regular basis. Further investigation is clearly needed.

Cookie blocking is a related issue. First party cookies are those set by the site the user is visiting, and they tend to be considered more trustworthy than third party cookies, issued by another site such as an advertising provider. Our preliminary figures suggest around 0.2% of visits are blocking first party cookies. Additional research is need to compare this to third party cookies, and to broaden the sample size. It’s not always possible to use a registration ID to increase the accuracy of web analytics, so it is essential that websites can understand the margin of error in cookie-based statistics.

Most web content is free to view, often funded by advertising placed against it. With the rise of the Firefox browser, which is easy to customise, there has been concern that more internet users would block advertising using plug-ins such as Ad Block Plus. Some media sites are blocking Firefox users altogether, but with ad blockers becoming more widespread for Internet Explorer, this will no longer be viable. Some sites suggest that 10-15% of their visitors use ad blocking, but our own preliminary figures suggest around 1.2% of visitors are blocking adverts. We would be interested in hearing from companies that would be willing to help with further research, including those that do not run advertising.

The final wonder of the web is tabbed browsing, which has become more popular with the growth of Firefox and the introduction of tabs in Internet Explorer. Opening content in tabs enables users to hop about the site more easily and read some content while other content loads. From an analytics point of view, it means the clickstream recorded can include illogical leaps between unlinked pages. By noticing differences between the page referrer and the previous page downloaded, it can be possible to tell whether the current page was requested in a tabbed session. Further research is needed to work out how many people use tabs, and what effect it has on how they navigate websites.

As I said, you’re invited to participate in this research. Data can be collected using WebAbacus, results will be fully anonymised, and traffic levels will remain confidential. If you’d like to learn more about what’s involved, please email me without obligation.