Interpreting Analytics – Things to Remember & Consider in Your Reports
Finding the most valuable data for your business within Analytics can be tedious work. You may also find yourself navigating in discouraging circles, enduring distractions that cause you to forget what you were analyzing, or completing the same tedious steps to generate the same report every month. Google Analytics has provided several basic tools to help eliminate these scenarios. I will outline these tools using the data for my own Leopard Nation website.
1. Visits Are Not the End All Be All
The first metric everyone wants to see when they open up Google Analytics is Visits. They want to see how many eyeballs are looking at their site. While this is a good thing to know, these visitors could be anybody and the real question we need to ask is “what are they doing when they come to the site?” The graph below shows a 169% increase in visits and a 291% increase in unique visitors. These increases look like a tremendous site improvement. But is it an improvement? Are these visitors the intended audience? Are these visitors engaging in the site? When we take a deeper look at the traffic coming into the site we see that there was far less engagement on the site than during the previous time period. There were 63% less pages viewed per visit and 74% less time spent per visit. Even more devastating was the 76% increase in bounce rate.
What can we take from this? Has the increase been a total disaster or has there actually been an increase in quality visits? We know that non-bouncing visitors are going to engage more than bouncing visitors by default. This is because bouncing visitors are all counted as spending 0:00 on the site as well as only viewing 1 page for their visit. The engagement metrics will obviously be higher for non-bouncing visitors.
Thus, in most scenarios we can conclude that non-bouncing visitors are higher quality, engaged visitors. Now when we compare non-bouncing visitors we get to see a truer form of quality visitor increase. For this scenario non-bouncing visits increased 7% overall and non-bouncing unique visitors increase 42% overall.
Thus the amount of quality visits to the site increased about 7% overall and the amount of quality unique visitors to the site increased 42% overall. These are much lower than the original increase in overall visits.
2. Remember the Long Tail
When looking at your top 10 keywords, referrers, etc. you have to remember that you are only seeing a very small picture of what is happening on your site.
The Top 10 shown here are just 0.009% of all the different variations of keywords and only 0.1% of all non-branded organic visits. Even when you segment your data, you can’t just look at the top keywords and expect the same performance throughout.
For example, in this segment we can see 8 out of the top 10 words had no new visits, but 41% of the entire segment was made up of new visits. In many cases it’s the keywords that only bring in a few visits that will make up a large portion of your segment. The industry name for these words is called the “Long Tail.”
112 out of 137 (82%) of all of the keywords in this segment only had 1 visit. These single visits accounted for 48% of all visits in the segment overall. Most visitors are going to type very specific keywords like these and that is why it is very important to optimize your site to catch as many Long Tail keywords as possible.
3. Have a Measuring Stick
You always want to compare your data to see what progress has been made. This may be done through comparing data in two different time periods, comparing data to the site average, or comparing the data to goals you have set up as a business.
Google Analytics makes it easy to compare your data to previous time periods. When viewing past data as your comparison there are three rules that you should always keep:
I. Equal Days
Both time periods that you are comparing should have an equal amount of days. While this may seem obvious, this is a common mistake among the analyst community. They will compare two different months with 30 days vs. 31 days or compare different quarters that have a different amount of days.
II. Equal Weekends
Another common mistake by analysts is not comparing equal weekdays and weekend days.
In the graph above you can see that the spikes in traffic are not lined up with one another and that the orange line representing September traffic had an extra weekend day. In the graph below the days are lined up correctly.
III. Equal Seasons
If your website is seasonal, make sure you are comparing time periods that have the same ideal type of traffic. You want to have equal sales cycles being measured or equal interest in the time periods being compared. Thus comparing year-over-year may be your best option for these cases.
For example, for my high school sports website there are no sports being played during the Summer, so it would not make sense for me to compare my traffic during Football season (my busiest month) to months in the Summer when there are no sporting events going on like in the graph above. But if I compare football season year-over-year you can see that there was a steady increase of traffic over all in the graph below. At the end of August 2011 I had a video on my site that went viral nationally and thus traffic on my site spiked. But other than that, the comparison is quite fair and the audience is similar.
4. Nothing Speaks Better than $$$
While comparing visits and other dashboard metrics can give you a general feel of visitor activity, nothing translates success better than comparing actual dollars and cents. This is why it is so important to create goals or setup E-Commerce in Google Analytics. With goals created and revenue attached, you can look at specific segments to determine which segment is making the most goal revenue or is converting the highest. Per Visit Goal Value is one of the best metrics to use in comparing the success of segments.
This way you can put more time and effort into expanding the successful campaigns/segments that have a small amount of visits and also look to make changes to the non-successful campaigns/segments that have a large amount of visits. This can also apply if you are using an E-Commerce site as you will be looking to find out which segments are selling and which segments are not doing so well.