Internet Librarian 2011
Understanding Users and Improving Your Website with Google Analytics
SuHui Ho and Jeff Wisniewski
Why use web metrics? The basic hit count is extremely misleading (we’ve known this for a long time). Looking deeper is essential. Some of the reports Google Analytics offers that are useful for us are the Top Content Report and the Traffic Source Report.
Top Content Report: The content on your website is a living being. So you have to keep it updated. We always have more content we can update. We have more rooms to clean on our websites than we have people to clean them. If you know which content is used most, you can prioritize those. Look at what the top tasks are too – what do people come to your website to do? Which pages from the homepage are clicked on most often?
Traffic Source Report: How much of your homepage traffic comes from the homepage settings on your in-library computers? SuHui says “don’t be so proud of your total hit count” because much of that direct traffic is from your own computers. J Referring sites are really useful to look at as well… And look at search engine traffic too – how much of your traffic comes from a search engine query result? Really pay attention to search engine optimization – make sure that your website appears on the top of the search results you want. Reviewing the keyword referrals from search engines is useful – you might see a lot of mentions of the word “hours” with variations of different phrases.
Jeff talked about the Goals and Funnels feature in Google Analytics. A Goal is the page a visitor reaches once (s)he has completed a task or an action…an end point of sorts. A Funnel is the optimized steps that the person should go through to get to that page. Looking at both of these in conjunction you can see where people get tripped up.
An example: the goal to register for a class on RefWorks. There may be several steps in your Funnel (e.g. look at calendar, click on class time you like, click on register, etc.). Per profile, you get to have up to 20 goals. Jeff says this is a reasonable limit. Name your goal something intuitive, because this is what shows up in the analytics reports. You can set the goal’s position, and pick a goal type (URL destination, time on site, pages/visit). For us, the URL destination is the most relevant to our type of web business. Be careful about leading or trailing slashes or spaces because Google Analytics will not match up the URL correctly. Once you set a goal, you get asked to create a Funnel. Then you set up the different steps—a Funnel can have an unlimited number of steps.
And then you wait to give Google Analytics time to collect data. Now that they’ve rolled out real time analytics, that’s awesome. You used to have to wait at least 24 hours, you can now see some data within minutes. With Jeff’s example, we see that we started with 886 people. 43 people didn’t get past the first step of the process, and it tells you where they went instead. The next step lost another 36 people, and then the 3rd step (checkout confirmation) resulted in a significant loss of 317 people. Whoa. That’s not good. But it shows you where you need to make changes and improvements, and in that way it’s super awesome useful! Whoopah! Jeff recommends looking at trend data over time instead of hanging your hat on a particular number at one specific time. Good advice!