Austin, Texas – http://www.localsurgemedia.net
One of the most important questions you should be asking yourself is which landing pages you should be focusing your promotional efforts on. Which pages are worth building links to, and emulating in your future efforts? This article will remain agnostic on which metric matters most (comments, subscriptions, time on site, pages per visit, sales). Instead, it will focus on how to interpret your analytics data.
This is not necessarily the most rigorous way to take advantage of analytics data, but it is a good balance between time invested and insight achieved. You may want to consider more scientific approaches that take statistical significance into account if you have the resources.
To take advantage of as much data as possible, you’re going to want to expand your date range:
- At the very top right section of the page, you will see your date range. Click the arrow on the right hand side.
- Change the year in the left-hand date to a date before your site was live, or at least before it was connected to Analytics.
Now you can start digging into the data:
- From Google Analytics, click on Content in the sidebar
- Click on Site Content
- Click on Landing Pages
- From the bottom of the page, click Show rows and select 500
- Above the results, click Secondary dimension
- Select Traffic Sources
- Select Source
Now, next to each piece of content, Analytics will share data from each different source of traffic. Since traffic from each type of source behaves differently, you will want to focus specifically on traffic from one source. This ensures that you are comparing apples with apples. Here’s how to do it:
- Above the results, click Advanced
- Click on Landing Page and change it to Source
- It should now say Include Source Containing followed by an empty text box
- Enter your desired source in the text box. For example, type “Google” into the box. Analytics will recommend sources when you start typing.
- Click the Apply button just above the results
Now your list consists of traffic results from only one source, so the data isn’t skewed in favor of pages that get more traffic from better sources. These are all landing pages, so the data isn’t skewed in favor of pages visited by people who are already browsing the site. Now you can focus on your goal metric. Here’s how you would do it for the number of pages per visit, so that you could figure out which landing pages causes visitors to view the most content:
- Click Pages / Visit to organize your content by the most pages visited by people who started on each landing page.
- Now your top pages are most likely flukes. They are likely single pages where only one or two visitors landed, and by chance had a high number of pages per visit.
- You will need to eliminate these flukes. Start by going back to the top of your list and clicking edit next to the field that says Advanced Filter ON.
- Click Add a dimension or metric
- Click Site Usage and then Visits
- It should now say Include Visits Greater Than, followed by an empty text box
- Here’s where things get a bit unscientific. Pick a reasonably high number that you think will eliminate the statistical flukes, like 10 or 100, and plug it into this text box
- Click Apply
Now you have eliminated most of the statistical flukes and you are left with the pages on your site that have the best performance history. From here you will want to use common sense to:
– Rule out pages that are no longer relevant and seem too dated to be worth promoting now
– Decide where overall traffic is more important than pages per visit, or whatever metric you are measuring. This is especially important when the differences are small. Bear in mind that 2.3 pages per visit isn’t much different from 2.6 pages per visit
– Decide which pages can reasonably be promoted any more than they already have been. They may already be sitting at the top of the search results, or the competition may be too rough for any improvements to be made.
– Consider revising the traffic count if your gut is telling you that the results still feel more like flukes than promising prospects.
Once you have identified your best landing pages, you can start running split tests on them to get even more out of them. You can also compare your best pages with your standard pages in order to improve overall site performance.