Website data has been used by analysts, designers, programmers and content creators for over 20 years to create better websites and get a better understanding of their users. During that time, the metrics we use to analyze our websites have gotten more and more sophisticated. But sometimes, we use the old standbys too much, use some metrics in the wrong situations, or use them as a single source of truth. Let’s look at some of these overrated or overused metrics!

Page views

I don’t think many people will be surprised to see this metric in this list. However, page views are used time and time again in situations that make almost no sense.

Page views is one of the few metrics used in digital analytics that is understood almost universally. In fact, Page views have been used since the dawn of World Wide Web. I even put a page view counter on my mid-90’s Geocities website.

Page views also involve big, impressive-looking numbers and are easy to boast about in any format. They are a commonly used advertising metric as well, further inflating their importance.

However, what do you really measure with page views? Some people might say “engagement” or “popularity”, but do we really know that someone engaged with the content on that page from the page view metric? Do we know how they feel about the page?


With page views, we have no understanding of intent, quality or engagement. All we know is that page was loaded. That’s it.

The user could have been lost because of poor navigation or pathing. Maybe they are looking for content that should be on that page but isn’t. The user might even be frustrated. They could be on your site accidentally. We have no further context—all we know is a page loaded. We need other metrics and tools to help us measure quality and content engagement.

Bounce Rate

Bounce rate is the metric everyone wants to use to evaluate the effectiveness of pages and content. In some cases, bounce rate can be used (along with other data and context) for evaluation. It’s a popular metric because it’s simple, used often used on dashboards, and is easily available in a quick roll-up on all analytics platforms.

Don’t get me wrong, bounce rate isn’t a bad metric. It’s just miscast as a stand-alone, one-stop shop for engagement.

Perhaps it would be good to quickly define bounce rate, as it’s often a metric that people think means different things. In most analytics tools, bounce rate is defined as a session that consists of just a single page with no other defined interaction, such as loading another page. Typically, there is no timing associated with bounce rate. To most people a high bounce rate means that page or site is BAD.

However, there are a lot of reasons your site or some pages might have a high bounce rate that don’t necessarily mean your site or page is bad.

Maybe your site’s homepage is set as the default start page on internal computer’s browsers and aren’t filtering internal traffic. This can also happen if it’s set as the default page after a user logs in to your wi-fi network.

Perhaps your landing pages provide excellent information, like a restaurant with its address or phone number on their home page. That information might be exactly why users usually visit that page or site. Once they get what they need, they leave the page.

It’s also dangerous to evaluate pages en masse based on their individual bounce rate. With the rise of intelligence in search engines, users get more and more effective landing pages that sometimes contain all of the information the user needs. Take a “Contact Us” page for example. I’d expect that page to have a high bounce rate, but I’d also suspect that page is doing its job!

Bounce rate is a metric that needs to be paired with other data and context. Combining bounce rate with other metrics and forms of measurement, like feedback and surveys, can be powerful. My advice is to ask your web audience about the quality and effectiveness of your pages and website directly.

Time on Page/Average Time on Page

This one is tough, because this metric seems straightforward. Sadly, most people using this metric haven’t looked deep enough into how this metric is calculated. There isn’t a magical gnome out there measuring every user’s time spent on a page with a stop watch. I know – I was sad to learn that too!

Just like bounce rate, a lot of people want to use this metric to measure engagement on a page. I’d like that too, but average time on page is not the metric everyone thinks it is. For most analytics tools to calculate this metric, there needs to be both a visit to the page that is being measured and then a page view on ANOTHER page with the same analytics script. Analytics JavaScript create time stamps upon a page load, but don’t have a way to timestamp upon exit. What this means is that an analytics tool has no idea how long a user was on a page if it’s the last page they visited during a session on your site.

Adding another wrinkle, time spent on a page or site doesn’t always translate to attention paid to that page or site. If a user gets distracted by an incoming e-mail, another site, or their 4th cup of coffee for the morning and is still on that page, all that time that has nothing to do with the page is counted until the user’s session times out.

With a full understanding of how time on page is calculated, it makes it hard to consider time on page as a reliable metric. You can look at adjusted versions of this metric that make more sense, but even then, it’s not a metric I use lightly. One way it can make some sense is to look at trending data, but I’d still make sure to put a large asterisk on any of those reports!

In my next blog, we’ll look on the other side of this web analytics equation and focus on underrated and under-used metrics!