Google Analytics — The Experiment

This blog is, in part, a venue for sharing my newfound Web analytics knowledge. Its primary purpose, however, is to provide me with an opportunity for hands-on learning with the tool Google Analytics. Once I added the tracking code to my site, Google Analytics began collecting data about visits to the blog. Now that it is time to look at the data, my first reaction is that I wish there were more to analyze! A brand new blog about a niche topic written by a non-expert does not exactly shoot to page 1 of the search engine results. As such, one interesting report to view is Traffic Sources, as it tells me how my few visitors are finding the blog.

About one-third of people who visit my blog were referred by another website. Looking at the Referral Traffic report tells me that all except one of these visits originated from Facebook, which is not surprising to me, as I had posted a plea on the social networking site, asking friends to check out my new blog. Interestingly, the report divides Facebook visitors into those coming from the full site (two-thirds) and those coming from the mobile site (one-third). As this report offers pages viewed per visit, visit duration and bounce rate, I can see that the full Facebook site sends better quality traffic. These visitors view 2.5 pages per visit (compared with 1.2), spend an average of 3:11 on the site (compared with 0:05) and bounce just 55% of the time (compared with 80%). These data suggest that people are less likely to read a blog entry or two when they are viewing the site from a smartphone.

The Mobile Overview report backs up this result, showing that visitors who are not using a mobile device spend an average of 4:27 and bounce only 47% of the time, while those using a smartphone or tablet spend just :34 and bounce 78% of the time. What Google Analytics cannot tell me, however, is why. Maybe those using their mobile devices to check Facebook were doing so “on the go,” and did not have time to peruse the website. Maybe they do not like to read on a tiny screen. Or maybe they read one entry (enjoying it immensely, of course), and then closed the window. In the latter case, the bounce is not necessarily a negative indication.

For an e-retailer, it is vital to figure out the “why” and then fix it. With more and more people searching the Web and conducting research on their phones, Internet-based businesses have to ensure that potential customers can access and browse their websites with whatever device they choose.

Earlier this year, research firms Sterling Research and SmithGeiger found that “two-thirds of smartphone users say a mobile-friendly site makes them more likely to buy a company’s product or service,” and “half say that even if they like a business, they’ll use its site less often if it doesn’t work well on their smartphone” (Hof, 2012, para. 4). Like with brick and mortar shopping, people want convenience with online shopping. Whether in a store or on the Web, if a consumer cannot find what she is looking for easily, she may leave to try one of the many other options available.

Back to my analytics…Complicating the data I am collecting is the layout of the blog, which features the three latest posts on the home page. My original thinking was that showing more entries on one page might encourage visitors to read more. I removed the barrier of clicking to a new page to see the next post. Now that I am looking at data, however, I realize that I also eliminated my ability to gauge how many posts are being read.

Fortunately, Google Analytics offers a method to account for bounces that may be positive, for example with blogs on which someone can accomplish a goal (reading a post) on just one page or with landing pages that have such an effective call to action that the visitor picks up the phone and calls the business instead of clicking through to more pages. One can designate a period of time spent on a page beyond which the visit would not be categorized as a bounce (Petrov, 2012). A blogger might specify 30 seconds or a minute, as her page likely has a significant amount of text, while a marketer might specify only 10-15 seconds if the landing page has minimal information. To set this adjusted bounce rate, “you should decide for yourself the amount of time you consider the user to be sufficiently engaged with your website or product” (Petrov, para. 7).

Another useful feature of Google Analytics is the Visitor Flow report, which offers a visual, like the one below, of how visitors navigate the website. Knowing how many people visited each page on a site is informative, but knowing the route they took to each page is much more valuable.

As my blog is quite new and has only four entries, there is only so much navigating a reader can do. Nevertheless, the Visitor Flow report shows the path visitors take from one page to another, as well as the page from which they exit the website. This report could be especially valuable for an e-commerce site that is designed to lead visitors on a path toward an online purchase, as a marketer could see where people are deviating from the desired path and then troubleshoot to determine how to make the path easier or more attractive to follow.

Wes Wilson (2012, “Visitor Flow”), CMO of Web hosting company ZippyKid, suggests that Visitor Flow charts are helpful for visualizing the big picture, but “despite their visual appeal and their rich display of information, they aren’t the most adept at providing conclusive, actionable metrics.” While I agree that a smart analyst is still needed to get to the reason behind problems, using segmentation with Visitor Flow can point the analyst in the right direction by illustrating issues that are specific to particular groups of visitors.

Perhaps shoppers outside a geographic area exit the site at the page where they find out the exorbitant cost of shipping the item to be purchased. Perhaps those using mobile devices exit at a page that is difficult to read on a smartphone. Perhaps users from a particular ad group exit on the product specification page because the item is not what they expected based on the ad. While Visitor Flow does not provide the answers, it can give some telling clues.

As I learn more about Web analytics, the enduring lesson seems to be that, regardless of the number of tools or amount of data you have, you still need a smart analyst to make sense of it all.

References

Hof, R. (2012, Sept. 25). Google research: No mobile site = lost customers. Forbes. Retrieved November 23, 2012 from http://www.forbes.com/sites/roberthof/2012/09/25/google-research-no-mobile-site-lost-customers/

Petrov, A. (2012, July 25). Tracking adjusted bounce rate In Google Analytics [Weblog post]. Retrieved November 23, 2012 from http://analytics.blogspot.com/2012/07/tracking-adjusted-bounce-rate-in-google.html

Wilson, W. (2012, July 30).  Tracking visitor behavior with Google Analytics [Weblog post]. Retrieved November 23, 2012 from https://www.zippykid.com/2012/07/30/tracking-visitor-behavior-with-google-analytics/