As I mentioned earlier, one of my main tasks for this year is to start looking at Search Engine Optimization for our sites and for the rest of campus. Part one of this was to learn about how searches work. Part two was to assess how at least the sites we maintain fair when it comes to search. Part three will be to make recommendations on how to improve them. Part four will be to extend this campus wide and hopefully help you improve your status as well. I am currently toward the end pf part two. I have compiled a preliminary document on the search rankings and optimization of our sites and have sent it to our department head. In order to share findings I will post sections of that report here as individual articles.
Please help with the project
I hope you follow along, and if you do so please consider doing the same analysis on your own sites and sharing that information with me. It would be very helpful to get an overall campus perspective since we don’t have much visibility into the numbers your sites receive. The information found here can be found in a combination of Google analytics, Google Webmaster Tools, and your login to the Google Custom Search (if you have set up your own account instead of pointing to ours.)
Search engine breakdown
Search engines accounted for 48.5% of all incoming sessions to www.tamu.edu. The breakdown of how much traffic each search engine generates is as follows (percentage is of all referrals, not just those generated by search engines.)
- Google – 40.5%
- Bing – 5.8%
- Yahoo – 3.0%
- Baidu – 0.65%
Google certainly is skewed because our local search is powered by Google, and we can’t separate out those from public searches.
Traffic generating search terms
Site-based Google analytics allows us to see which terms people used when they clicked through to our site. An analysis of several sites – www.tamu.edu, m.tamu.edu, tamutimes.tamu.edu, visit.tamu.edu, and experts.tamu.edu – shows that the overwhelming majority (65-75%) of all referals come from “not provided” as the key work. A little research indicated that is an indication of people browsing while logged in to Chrome and/or other Google services. When people browse in a logged-in condition Google will not pass along their search query information. The rest of the top ten key words tend to be permutations of “texas a&m university.” Content level key words tend to make up no more than 0.5% each.
The content terms that do turn up tend to be site specific – information about jobs and university location/address for the university site, article topics for TAMUtimes, expert names and categories for the experts list.
If we assume that the “not provided” search terms are substantively the same as the terms what we can see, we can make some assumptions about user behavior. This analysis indicates that the people who are coming to our pages are already looking for content that we provide. That could be considered a good thing – they are getting the content that they are looking for – but it is not being as effective in casting a bigger net for those looking for a more general topic and pulling them in to our sites.
No comments yet.