Google Analytics (GA) use is as widespread as it is useful. Unfortunately, the default results can often be misleading to the uninformed Analytics user. I'll share one method to filter out a lot of the "static" in GA by clicking a single checkbox within the recently-introduced Advanced Segments feature...
Google Analytics: Basic Stats and "Static"
Once installed, Analytics seems almost magical to a first-time stats tool user and especially magical to DotNetnuke users who have relied only on the Site Log to inform them about their visitors. With Analytics you suddently you see far more than just page hits and source URLs. Average time on site, average number of pages viewed per visit, top "exit" pages (pages that "lose" visitors), browsers, screen resolutions, reffering sites, visitor loyalty, top search keywords, connection speeds....whew! In short, tons of great info all for free.
Unfortunately, many of the same first-time stats users who are so enamored of the new info may also be mislead by it. See that stat that says "87% bounce rate"? That's the number of visitors who left after viewing just one page. So what? "They found what they were looking for and then left" you might think. The typical trend reveals something very different. Typically, the bounce rate correlates very closely to the percentage of visitors who did not find the content they were looking for and, therefore, are probably not your intended audience. Such data amounts to "static" that prevents you from drilling down to data pertaining to your target audience. That's a good point to be aware of but how can you tell for sure what is "static"? If you do have "static", how do you get rid of it?
GA: "Bounce Rate" and your True Audience
Log into your Analytics account. Make a note of your "bounce rate". Now use the menu to navigate to:
VISITORS -> VISITOR LOYALTY -> LENGTH OF VISIT
We can safely assume that anyone who leaves a page within 10 seconds of loading it did not find the content they were after.
Compare your "bounce rate" stat to the "0 to 10 seconds" length of visit stat. Unless you're running a single-word spell-checker site, the two stats will probably be very close to each other. Further, the "0 to 10 seconds" stat will likely be far, far greater than any of the "more than 10 seconds" stats. Perhaps many people started with your homepage and immediately clicked through to a different page (like visitng "DotNetNuke.com" and immediately clicking the "Forums" link). That's possible but is unlikely for sites without high visitor loyalty (visitors already familiar with your site). More typically, the "bounce rate" is an indicator of inappropraite visits -- visits by people with no interest in your content. Such visitors are not your intended audience and amount to "static" when you're trying to gleam data about the visitors who do spend time on your site.
This distinction is important because if your "bounce rate" is, say, 87%, then that means that up to 87% of the rest of your stats may not reflect of your true target audience. You want to know about the people that STAY on your site so you can judge what's hot and what's not. You want to be able to tailor content, style and layout based on your target audience.
Examples:
- Do you still need to use a skin that looks good in 1024x768 or has most of your target audience graduated to 1280x1024?
- That cool new module really takes your site to the next level but.....it doesn't work with Safari. How many of the people that spend time on your site are also using Safari?
We need to filter out those "bounce" visitors ("static") and drill down to the people who actually stay on your site.
GA: Advanced Segments
Introduced in October, 2008, Analytics now has just the option we need: Advanced Segments (AS). We can use AS to segment (filter) our data into different catergories. Even better, our different category stats will be presented side-by-side for easy comparison. Let's go ahead and filter out those "bounce" bums...
In the upper-right corner of GA, find and click the "Advanced Segments" dropdown box. Now under the "Default Segments" box that appears, scroll down and check "Non-bounce visits". Leave the "All Visits" box checked. Click the "Apply" button and you're done.
Now all of our stats and graphs will show TWO sets of data: the original "all visitors" data and a new "non-bounce visitors" set of data. Go back to that "length of visit" page (Visitors -> Visitor Loyalty -> Length of Visit). Now you can see a much more accurate view of how long your true target audience is staying on your site.
Caveats -- Many ways to interpret data
I've made the broad assumption that someone who leaves your site within 10 seconds did so because you did not have the expected content and, therefore, that perosn was never part of your intended audience to begin with. This is a reasonable but also quite fallable assumption. Perhaps the person IS your intended audience but your site does such a poor job of presenting content that the person leaves without realizing the sought-after content is there. Perhaps you run a spam-ad site (so everyone is your intended audience) but the person leaves as soon as its apparent that you're not really offering credit protection services despite your AdWords ad.
Another false implication is that data with "static" will vary from data without "static". For many stats you will probably see a direct correlation between the two sets of data. For instance, trends in "browser capabilities" stats will probably be very similar. Conversely, "visitor loyalty" stats will likely be quite different. Thus, removing "static" is not a cure-all. It will be much more useful for some stats than for others.
Summary
Suffice to say, even sites with the best intentions and appropriate content will likely have at least some "targeted" visitors accounting for the site's bounce rate despite being the site's intended audience and vice versa. In general, analytic data should be viewed from afar -- it's much better at revealing "trends" rather than "absolute certainties". Learning how to fine-tune your reporting tools and filter out misleading "static" will aid in identifying those trends. GA's Advanced Segments is a fantastic feature for helping Analytics users do just that.
Additional Reading: Advanced Segments
We only covered the tip of the iceberg in this post. Watch and read about other ways to leverage Advanced Segments...
Intro to Advanced Segments: A Video
http://www.youtube.com/watch?v=wu8YzF0AM14&feature=PlayList&p=AFDC0271A9E3C759&index=0
Example: Create a custom Advanced Segment
http://analytics.blogspot.com/2009/02/urban-apparel-and-advanced-segments.html
Advanced Segments: How many of your visits are from iPhones?
http://analytics.blogspot.com/2009/02/how-many-of-your-visits-are-from.html