Finding Answers in Google Analytics

Google Analytics contains enormous amounts of information, yet many website owners leave a session with their analytics knowing no more than they did when they started.

How can you make sure to gain actionable information from your next visit to Google Analytics?

Ask a question.

While you can often find interesting things by poking around in your analytics to see what you can see, it’s more efficient to go in with a question in mind. Here are some examples of questions you can answer with analytics:

  • How do our customers find us?
  • Is Facebook or LinkedIn more likely to send us converting customers?
  • What kinds of links send us the most engaged visitors?
  • Where in the checkout process are we most likely to lose people?
  • What geographic area holds the customers most likely to buy our product?
  • What’s our average placement on the search engine results page for the keywords that bring us visitors?
  • What kinds of information are most interesting to our visitors?
  • Which sales landing pages perform best?

… and many, many more. Note that all the questions here are about our visitors’ online behavior. Sometimes we can extrapolate to other aspects of our relationship with our customers, but there are definite limits to what we can find out directly.

Identify the dimension and the metrics.

If we want to know which blog posts are most engaging for our readers, we should look at the Behavior reports and consider the length of people’s visits or the number of pages they read in a session. If we want to know which geographic area to target, we can compare conversion rates among customers from different locations.

Neither of those questions can be answered by looking at our traffic.

Basically, this step involves identifying the reports in which we’ll find the answer to our question, and the things we should be counting or calculating to get that answer. Often, we can come at the data from more than one direction. In such cases, it makes sense to look at both possibilities, to see what information each can provide.

Segment the information.

While this step may not always be necessary, it usually allows us to dive more deeply into the data. In the screenshot below from our lab site, we are looking at our referral traffic, but only at visitors who convert. As we refine our question and look at different aspects of the data, we will always be looking just at that segment of our audience. That eliminates a lot of the noise and lets us catch the signal more clearly.


Connect with the real world.

Often we find that we need some information about the world our visitors live in before we can fully answer our question. Were items out of stock this week? Were there surprises in the weather that might affect our numbers? Did political events or a holiday distract our target customer? When we look at the big picture — quarterly changes in traffic, for example — we can ignore these fluctuations, because we have enough data to smooth out these small differences.

When we’re looking at a more granular level, we often need to consider the things that are going on in the real world. One client recently saw  a drop in conversion rate for one type of sale: class registrations. As we were tracking down the causes, we discovered that many of the upcoming classes were full. People who wanted to register for classes were not able to — something Google Analytics can’t measure.

Following these steps can help you find the information you need in your analytics reports and get more value from them.







Leave a Reply