The Content Authority Blog

This Blog Is Written By The Content Authority

In our previous post, we talked about mining data from Google Analytics. Sometimes it’s hard to know what to do next once you have gleaned data from your website. As web marketers with access to loads of useful data, we still spend much of our time guessing about exactly what will please our visitors. This post will offer some tips on what to do with data once you are done analyzing it.

Engaging Your Visitors

Forming deep connections with visitors to your website is an important part of building a brand online. Identify pages where users are spending the most time. What is it about those pages that keeps people on your site? Mimic these characteristics on other parts of your website. You can also look at your keywords report to see which keywords are correlated with long stays on your website. This clues you in to similar key terms that may be useful to target on other pages of your site in a search marketing campaign.

Top Entrance Pages from Search Referrals

Take a look at Traffic Sources -> Search -> Overview, and then add a secondary metric of landing pages. These are the pages that visitors are finding and clicking on in searches. Your most visited page may be your home page (which is fairly common), but take a look at other pages that are ranked highly in the number of visits. These pages are doing very well in search engine results pages, and you can compare them to pages that may not be showing up so high in the reports to help you improve their rankings.

Real Time

At first, the real time feature in Analytics is cool to look at, but you might consider it to be a novel addition, at best, after a short time. There are actually a couple of cool uses for this feature. It’s not wise to make programming or content changes to a live site, but it can happen sometimes. You can use the real-time report to see if anyone is looking at the page you plan on changing before you take action. Other uses for the report include checking initial progress of a campaign like an email blast, sharing content on social media that involves a link to your site or any other activities that involve sending traffic to your property. You should always have some other, more recordable method of tracking for a campaign, but the real-time report can give you a good initial picture of how your campaign is shaping up.

Site Speed

Having a slow website can increase bounce rates and, in general, promote a poor user experience. The site speed report is one that gets little fanfare, but it’s pretty useful. You can see the speed of your site’s average load times in most major browsers and mobile operating systems. You can also look at load times on the page level. If you see that load times are too high, you can take proper action. Google provides a site speed analysis for free, and it will tell you exactly what may be causing your site to move slowly.

Observe, Hypothesize, Test, Repeat

In any post about how to use data in Analytics, testing should always be mentioned. Sometimes there are small tweaks that can be made to a website, but often our assumptions must be validated by a series of tests to ensure that we are making the right decisions.

A/B/n and Multivariate Testing

After data has been analyzed and conclusions drawn, don’t just start changing things around on your website or individual pages. Not only will this confuse you down the road, but you will not be able to determine if your educated guess about what to change was, in fact, correct. Instead, you can set up carefully administered tests. Two of the most common forms are A/B/n tests (also known as A/B testing or split testing) and multivariate testing.

A/B/n testing

  • This is a very common form of testing used in web design and web usability. It involves making two versions of a page and sometimes three. Traffic is then divided among the pages to see which variation has the best results in terms of a goal. The goal could vary, and some examples might be increasing time one page, increasing conversions, decreasing bounce rate or getting visitors to perform some kind of action. The benefit to this type of testing is that it’s easy to set up and implement.

Multivariate testing

  • This is similar to A/B testing; however, it involves testing multiple variables on a webpage. Another way to think of this kind of testing is many A/B tests happening on the same page at the same time. The main purpose for this kind of testing is to determine which combination of variables produces the desired result based on the goal of the test. Multivariate tests can be very difficult, time consuming and expensive to undertake. It is best to start out with A/B/n testing first, especially for smaller tasks.

Testing your changes is very important. Keep a log of how each variation of a page performed when you set up your tests. For an A/B test, narrow down your best guesses as to how you could better accomplish your goal with the layout and elements of a page. Then, pick the two best configurations (or three if you have that many), and start your test. Here is a great resource on setting up an A/B test for your website.

What you do with your data will largely depend on what it is telling you. High bounce rates, low visit durations, low numbers of pages viewed and scant referrals or conversions will all have their own solutions. Make sure you are analyzing your data with the context in mind. For instance, a high bounce rate on a single page blog may not be an indication of copy that isn’t engaging or interesting. Low conversion rates could be attributed to other factors like a long or confusing sales process.

What decisions have you made for your website based on Analytics data? Did you find that further testing helped identify what you needed to change on your website?

About Shawn Manaher

Shawn Manaher has written 384 post in this blog.


Loading Facebook Comments ...
There are no comments.

Leave a Reply

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>

End Comment -->
Navigation