Skip to Main Content
Close Menu

What Metrics to Check Before a Redesign

What should you consider from an analytics perspective when prepping for a redesign?

Illustrated collage of website statistics

In my first Web job a decade ago, I was a news producer and managed the overnight updates of a site. As I picked which stories went where and edited features on the site, I was sure all our visitors used it just the same way I did. In my mind, I was the typical user.

The day I got my first access to reports on how users were actually using the site, my mind was blown. Yes, my patterns matched a subset of users, but it was a small subset — okay, a really small subset. Suddenly, I realized why our site was evolving some of its features the way it was. I also had a clear idea of where I needed to spend more time to really give readers a better experience.

In the years since, I've tried to keep that experience in mind whenever I've had to make decisions about where my priorities should be -- especially during major or minor redesigns. Sure, I could pick the option that helps me the most in my daily view or focus on something that I think would be the best. But, like I found out a while ago, I just might not be the typical user for a particular instance.

If you open a Google Analytics report up, you could spend hours clicking around. It's kind of fun in a nerdy way, I'm not going to lie. But here's the thing: With so many numbers, rates, and conversion numbers possible, what should you look at first to start making decisions for goals?

The answer comes down to what matters most to you. A site with dozens of employees may have hundreds of numbers to evaluate, but there are a few you simply can't overlook. Be warned: once you start looking at the numbers, you're bound to become a little obsessed.

How are people getting to your content?

Finding out the main ways people are getting to a site -- bookmarked homepage, search, social, and peer-to-peer sharing -- is my first stop in checking out audience behavior. You might have the most gorgeous homepage in the world, but if everyone just finds you via social postings, geeking out that homepage during a redesign may become less important.

Typically, this data will be under Traffic Sources in your tracking program. Traffic sources typically can broken down to:

  • Direct: Traffic that goes to you by bookmarks or someone typing in your URL. You also can look at Referral traffic to see if a particular site is giving you a lot of link love. Sometimes this is also people sharing by text messages or in instant-messaging clients.
  • Social: Traffic from Facebook, Twitter, Pinterest, or Reddit. These users thought your content was so awesome, they wanted to share it with their friends.
  • Search: People using Google, Yahoo, or Bing. The user asked the Internet a question, and you came up as a result.

Once you find trends here, you’re bound to want to dig more into each. For example, if you find you're huge on Pinterest, it would be worth digging more to find out how and why. Is there a particular type of post you’re doing that makes people find your content and want to share it for you? At the same time, if you're getting no traffic from Google, it’s probably worth pausing for a moment to see what you can do to improve that.

What content is being looked at?

If you have a Web site, you probably spend a lot of time on particular pieces of it. Are you getting the right amount of traffic to those items? Are people looking at what you want them to be looking at?

For those questions, you want to look at your page views. Page views tell you the number of times a particular page was seen in a browser. A person could go to the same page a few times and increase that number, so keep that in mind.

Here’s some conversation starting points for page views:

  • What is No. 1 on the list? Why? Can this page be improved to meet your site’s goals?
  • Do you see huge numbers on a small handful of pages and then negligible numbers on the rest? Are you spending time on the negligibly trafficked pages? Why do some of those pages have such low traffic?
  • Are there common threads going through your top items? Did you intend for that? Can you improve upon that?
  • Take a look at the items you expect to have high page views, but don’t. Why are your users not getting to this (display, deep linking, social strategy)?

What is turning people off?

Inevitably, people are going to leave your site. Unless you managed to allow shopping, news, and, you know, other... ahem... popular content all under your site’s umbrella, people are going to want content you’re just not going to be able to offer. But, a key to site improvement is to know from where people are leaving.

There are metrics that will give you this and they go together quite nicely:

  • Bounce rate: The number of times a visitor first entered your site on a particular page and then went away. Think of someone getting your page in a Google result, clicking on your site, and then quickly going back to Google. If a page has a tendency to be a one-and-done, you might a problem. (Note: Check the time spent on these pages, too. If a page has a high bounce rate, but a high time on site, it could mean people got everything needed.)
  • Exit rate: These are pages that users left your site the most from. If this is a “Thank you” page after buying a product, this might be expected. But, if it’s the first page in the buying funnel, you might need to take a minute to revisit what’s going on. All Bounces are exits, but not all exits are bounces.

Once you get a chance to collect these stats, the next question to ask yourself is: “Why are people leaving?” If the page is designed to move people through a process, then maybe the process needs to be refined, or maybe the action for making the next move is confusing. Worse yet, maybe you have a broken link and people are getting aggravated.

After you find trends, you can then use your redesign as the opportunity to enhance or fix the problems your metrics pointed out.

What are people using to look at my my site?

Six years ago, I was finally able to say that Internet Explorer 6 was no longer a thing I needed to care about. The site I was working on had a huge audience that wasn’t able to upgrade its computers, so to keep that audience, we had to keep supporting that one browser. It added extra hours and thought to every move we made, and it added time to all our following testing.

Yet, we knew we had to do to it since our analytics system showed a nice pie chart of people using that browser. To really know your users, you need to know what technology they’re using to see your work. You can generally find this work under Visitor Profile in your tracking system.

Questions to consider:

  • How many users are on desktop, mobile, or tablet devices? If you have a lot of a mobile users, are you optimizing for them? If you have no mobile users, do you know why?
  • What browsers are being used? There are two things you want to get out of this area: What do you need to support? Do you look good in all the browsers you want to support. Generally, a site will look to support any browser with more than 5 percent of users.
  • Also, as a content editor, spend some time in the browser of your user. What do you think of your experience?

The more time you spend with your numbers, the more questions you’ll find that will lead you to a better site. Also, the more time you spend with the numbers, the more you’ll be able to tell the story of your users and learn how to make their experiences better.

Let's work together.

Have an exceptional idea? Let's talk and see how we can help.
New Call-to-action