Gender and Google searching

Link: google.pdf (application/pdf Object). is a very interesting report that I picked up via SEO by the sea. It's a paper from researchers at Cornell and Stanford who did the wiring up of eyeballs to see what people were doing when they were searching Google, and they came up with some interesting information. It's quite a long paper, and to be honest I skimmed most of it, but there are some points that should be brought out, which either I've done, SEO by the sea has, or the original authors did - take your pick!

• The query results abstracts are viewed in the order of their ranking in only about 20% of cases, and only an average of 3 abstracts per page are viewed at all.
• In 96% of cases only the first page of results were viewed, and in no case did any of the searchers move beyond the third results page.
• Users rely heavily on the abstract to decide if they will view a page result or not.
• In over 50% of cases when users revised the search strategy they did so without looking at any results.
• Users did not necessarily view results sequentially; 34% of scan paths were linear and 19% strictly so.
• Users tended to look at the first 2 results, and if these did not give the desired information they would skip around.

Consequently, it's important to be in the first 2 results, but if you can't manage that (from a website publishers perspective), anywhere on the first page is good. That's kinda obvious really, but worth repeating. If you're not in the first 30 results, well you just don't exist.

A few things were gender related - women submitted longer queries than men, with 3.8 words and 3.4 words respectively. This is interesting in that previous work has shown that people used to limit searching to about 3 words per search, but that seems to be increasing slightly. Men tended to look at the last 3 results on a page more than women, and were also more likely to view additional Google resutls pages (5.4 times in fact). Clearly therefore if you're selling products to women, you really do have to be on the first page!

Is any of this actually significant? The subjects used for the study were undergraduate students, and they only used a total of 23, with 400 queries and 600 results pages. So, nothing to lose sleep over I don't think, but interesting none the less.