How To Split An Atom

Get more articles like this and Register for our Forum.

Analyzing traffic can be a daunting task. There are so many ways to look at the data presented by the major analytics engines (Sitemeter, Google Analytics) that it can be daunting to separate what really matters from the surrounding noise.

This guide will introduce you to some of the most common factors to use when measuring a web applications success and help you interpret them.

Unique Visitors

How many real people are coming to your site?

You don’t need any deep knowledge of statistics to understand that the higher this number is the better it is for your product. From a marketing perspective, having a high number of uniques is one measure that your product resonates with a wide audience.

Since this is the simplest metric of site success, I will save discussion of its subtleties until we move to . . .

Pageviews

Pageviews give you the answer to the question, “When visitors come to my site, what do they do?”

Do they view your homepage and immediately leave, or are they pouring over dozens of sections? This is one of the major statistics that advertisers use to determine the worth of your site and for good reason. It is the metric that shows how deeply interested the audience you attract is to your content.

Pageviews relate directly to uniques, and when analyzing these two metrics you should keep two things in mind -

A high pageview to uniques ratio is a sign of a sticky site, one where users go and explore. People are clicking around. This is a great thing, unless the total number of uniques is so low that your own exploration of your site skews this ratio.

Having a low pageview to unique ratio is not necessarily a bad thing, especially if you are a content producer. Many times, even loyal users only come by to see one particular article or watch one particular video. Know, however, that this can also be a sign of a user base that isn’t sufficiently engaged with a product.

A good secondary check is . . .

Time On Site

How long are users spending on your site?

Advertisers don’t currently give much merit to this factor, but with the increased prevalence of flash applications and AJAX, it is becoming an important tool for developers to measure their relevancy.

One way to look at it is as a replacement for pageviews for a site that, by its nature, doesn’t generate pageviews (because it’s mostly AJAX for instance). This tells you how much time people are spending looking at your site. Unfortunately, it can also tell you how much time they are spending leaving your site open while running an errand.

At low levels of traffic, this metric can be completely skewed by development time (I am sure you spend a disproportionate about of time on your domain) but as traffic increases it becomes an increasingly good yard stick of how well your content is holding user attention.

A special note for content producers. As your traffic increases, many times your time on site will decrease. The idea is that there is a lot of “drive-by” traffic on the web, especially for platforms like blogs. All of the people visiting your site to see one thing and leave will, as your traffic grows, work to drive down your overall time on site. You should always be working to make people want to stay longer, but don’t be surprised when your next traffic spike sends this number way down.

This number is directly related to . . .

Bounce Rate

How many of the people that come to your site immediately leave?

Having a high bounce rate in conjunction with a low time on site is almost always a bad thing. What it shows is that people are showing up to your site, not viewing any pages and the one page they are viewing they aren’t spending much time on.

Keep careful watch of your bounce rate, and using other factors (referral statistics, for example) try to figure out where this “junk traffic” is coming from and why they are leaving. There is no point in having users come to your site if they don’t engage with any of your content.

Another related note is that a high bounce rate often means effects your . . .

Conversions

This is the last item in the chain but for retail-centric services, it is the most important. How many of the people coming to your site are actually making a purchase? If you don’t monetize your eyeballs through advertising, this number is much more important to you than any of the other metrics measured. It doesn’t matter how much traffic you can drive, if that traffic isn’t converting then what’s the point?

What do you think is the most important measure of a web applications success? Does it go beyond the statistics? If so, how do you measure it?

(Image) (RSS)