By Andy Crestodina published October 17, 2013

3 Content Optimization Questions That Google Analytics Can Answer

google analytics on laptopIt’s a big tool with lots of reports. Google Analytics can be intimidating. It can also be a huge waste of time. Too many marketers just browse through charts and graphs, without gaining any insights, without making a decision.

So rather than jumping into an ocean of data and swimming around aimlessly, let’s use Analytics to do our analysis for us.

Here’s what we’re going to do…

  • Ask a question about our website or visitors
  • Log into Google Analytics and find the answer
  • Make a decision and take action!

This action-oriented approach to Analytics is the key to content optimization. If you’re part of a big team, this approach will help build support. If you’re a service provider, this approach may help prove ROI. Either way, it’s the key to making good decisions.

Here are two examples of how to ask nicely so Analytics will tell us exactly what to do:

Which posts are most popular?

A study by InboundWriter shows that 20 percent of web pages drive 90 percent of traffic. For most sites, these are the pages that rank in Google, but they may also be the pages that are getting shared. Regardless, let’s ask Analytics and find which of our blog posts are traffic magnets.

Note: This example uses the Orbit blog, which is in a directory, orbitmedia.com/blog/. If your blog is in a separate directory or subdomain, filtering to see just the blog posts is a bit easier.

  1. Go to Behavior > Site Content > All Pages.
  2. Set the filter to show only blog posts (pages that start with /blog/).
  3. Set the secondary dimension to “Medium” (Note: This step is optional, but it will show you if visitors came from search engines (organic), other sites (referral), email campaigns, etc.).  
  4. Sort by “Unique Pageviews.”

sort by unique pageviews

The report you’re looking at may confirm the findings of the InboundWriter study — a few posts are likely driving a ton of your traffic.

If you’re going to spend 10 minutes working on your website today, work on the pages that are getting seen. Here are five ways to give these pages some love:

  • Make it more interesting with a great image, quotes, clear formatting, and concise writing.
  • Make it more informative with evidence, statistics, details, and examples.
  • Add a compelling call to action.
  • Add “related links” to other posts on similar topics to the bottom of the page.
  • Add internal links within the body copy to other pages.

Now we can gently guide visitors from these pages to our most effective pages. But which pages are the most effective? Let’s ask our Analytics…

Which posts are most compelling?

The posts that bring in the visits aren’t always the pages that turn visitors into leads, subscribers, and customers. In other words, traffic magnets aren’t usually conversion magnets.

As long as your goals are set up properly, you can check this for any type of conversion: lead generation, e-commerce customers, event registrants, job candidates, etc. Here’s an example of how to use it to find which blog posts are turning your blog visitors into newsletter subscribers.

1. Set the date range for one year. Especially for lower traffic sites, we’ll need a lot of data for this one.

2. Go to Conversions > Goals > Reverse Goal Path.

3. Set the filter to show only blog posts as “Goal Previous Step – 1.”

goal previous step 1

4. Sort by “Goal Completions”

sort by goal completions

Note: On our company blog, we include a sign-up option in every post. This is important because it allows me to set “Goal Previous Step” to be just blog posts. In other words, I can ask Analytics which post visitors were reading just before they arrived at the email subscribe thank you page…

Now you’re looking at a list of blog posts, along with the number of newsletter subscribers each has generated. But remember, the most compelling aren’t always the most visited. So really, we’re looking for the conversion rate for each post, not the total conversions. So there are a few more steps…

5. Go to Behavior > Site Content > All Content.

6. Filter this report so only blog posts appear.

7. Sort by “Unique Pageviews.”

8. Create a spreadsheet of these blog posts, with one column for “Conversions” and another for “Unique Pageviews.”

9. In the spreadsheet, divide the Conversions number by the Unique Pageviews number.

unique pageviews

Now you’re looking at the visitor-to-newsletter subscriber rate for each of your blog posts. In this example, you can see that some posts on the Orbit blog have been far more compelling to our readers than others (1.5 percent compared to 0.3 percent). That’s a big difference — some posts are literally five times as compelling as others.

post conversions

If you’re going to spend 10 minutes on efforts to drive more traffic today, these are the posts to focus your efforts around, as they are clearly the most compelling to your audience. Here are five ways to drive a little traffic their way:

  • Share them again on social networks.
  • Add a link to these posts in your email signature.
  • Write a roundup of your top posts and put these at the top.
  • Write a guest post on a similar topic and link back to the posts.
  • Make links to these posts more prominent in your navigation.

Analytics is more than a scorecard

Sadly, many content marketers don’t use Analytics to do analysis. They just check traffic, smile or frown, then move on with their day. But great content marketers are curious. They ask questions, form a hypothesis, test it, and then act on the evidence. Here are some other questions we might ask Google Analytics to answer for us:

  • Did simplifying our contact form increase leads?
  • Does our new responsive website connect better with mobile visitors?
  • Which social network is driving the most leads?
  • Which guest bloggers are creating content that gets shared the most?
  • Which of our blog categories or website sections are most popular with visitors?

That last question is really interesting. Many content marketers produce content within categories and sections without knowledge of what visitors are reading. The gap between Percent Created vs. Percent Consumed is easy to find and easy to fill. (Perhaps I’ll walk you through it in a follow-up post!)

What questions is Analytics answering for you? Share your experience with us in the comments.

For more guidance on using analytics tools to determine content optimization priorities, read CMI’s eGuide on Measuring Content Marketing Success.

Author: Andy Crestodina

Andy Crestodina is the Strategic Director of Orbit Media, a web design company in Chicago. Andy is also an instructor for the Content Marketing Institute Online Training and Certification program. You can find Andy on and Twitter.

Other posts by Andy Crestodina

  • Jake Parent

    Great post Andy. Did you see that Google is now offering a free course on analytics? Check it out here: https://analyticsacademy.withgoogle.com/

    • Andy Crestodina

      Yes, I saw this. I think this replaces Conversion University which was the home of the “GAIQ” test, aka the “Google Analytics Individual Qualification”

      Assuming it’s the same thing, I highly recommend going through the videos and taking the test. It only takes a day or so and you learn a lot about setup and options. It also looks great on a resume or in a LinkedIn profile… :)

  • http://www.docalytics.com/ Steve Peck

    Really appreciate the step by step guidance around making the information Actionable. My company offers a similar solution to Google Analytics, only surrounding downloadable online documents like Whitepapers, eBooks, Case Studies, etc.

    All too often we also see instances where content marketers don’t “Use analtyics to do analysis”, so are continually pushing our customers to take a more scientific approach and establish hypothesis around how their customers will engage with particular documents based on the content, formatting, and target distribution channel and then where necessary make specific changes to how the information is presented to see the impact on exactly how readers engage.

    When studied in this way (and as Andy points out so well in this piece) many marketers realize that the whitepapers they think are performing well as they are downloaded often, aren’t always the one securing deep engagement by the reader.

    Definitely looking forward to your follow-up on “Percent Created vs. Percent Consumed” as that is a key metric for us in evaluating content effectiveness.

    • Andy Crestodina

      Thanks for the input, Steve. We’ll be sure to let you know if/when we post the followup. Until then, you can always find great Analytics techniques on Avinash Kaushik’s website, Occam’s Razor. I’m sure you’re familiar with his content. :)

  • Charles Warnock

    What a great post. It’s a lot more structured approach to answering your (or your client’s) questions than just looking at popular posts and trying to produce content that’s similar to what’s already working. Forming a hypothesis is one of the most neglected steps in the process. The steps presented are great for those just learning and to keep top of mind for content producers who have been around a while.

    • Andy Crestodina

      Agreed. Thanks for the feedback, Charles. To me, it all seems like more of a science than an art. My gut might tell me something, but I don’t make decisions until verifying my gut with Analytics.

      I once heard someone say that the plural of anecdote is data. Make sense, right?

  • http://twitter.com/nicowegher Nico Wegher

    Andy, thanks for such a great lesson!

  • http://gitgrow.com/ Santosh Rajan

    Helpful post Andy.

    I agree with you and Charles that there should be a structured approach to Analyzing web data (for both one’s self or that of one’s clients).

    I will be adopting your methodology in the future.

    Looking forward to reading more from you.

  • Dianne Hazeltine

    Good article – check your link to inboundwriter

    • http://www.contentmarketinginstitute.com/ Michele Linn

      Thanks for the heads up, Dianne. The link has been fixed.