Potential Drawbacks to “Time-On-Site”

By Dave McAnally, Natural Search Supervisor, Content Solutions

Another couple issues have come about in the past couple weeks that prompted me to wax poetic on a subject (I hope) is at the forefront of any web marketer’s mind.

We talk a lot about query intent and site intentionality around here. Websites have a purpose. Some produce sales, some entertain/engage your user, some gather information, and so on. I think most people have a sense of what the key performance indicators are of a revenue producing site. There are many other non-monetary KPI’s out there used (phone calls, email clicks, brochure downloads, list signups, etc. etc. ad infinitum). One aspect of measuring a marketing campaign’s effort is the engagement it produces. A metric one can use to measure engagement is ‘time-on-site’.

I think time-on-site is an oft overlooked metric. At its most basic level, it tells you whether or not people hang around on your site or leave immediately. It can certainly speak to how your website manages to maintain your visitor’s attention. However, it must be taken with a grain of salt. Just because people are lingering around on the site longer than they used to doesn’t mean you’ve got a good thing going.

This brings me to my current situation. Picture, if you will, a website built by a brand with a TON of on and offline equity. This site, crafted by the finest web producers in the land filled with dazzling pictures, celebrity spokespeople prominently displayed and multimedia extravaganzas that are, oh, let’s use the word ‘shimmering’. The site is aesthetically in line with offline efforts and has multiple interactive features that a user can poke and prod for hours. There’s a basic HTML counterpart so folks who don’t have flash, or surf in on a mobile browser can get to the content. It’s a vast improvement over its predecessor in many ways to be certain.

So the day comes that the site launches and sure enough, analytics all point to the time-on-site shooting straight up. Furthermore, you see a noticeable spike in traffic because you are now ranking in places you weren’t previously and your new site is generating buzz in social networks prompting new visitors. All seems to be flowing smoothly except for one thing- the direct response elements of the page and key components of the page suddenly see a drop in activity with respect to the increase in traffic. The store locator or the email signup numbers remain the same or even start to decline. What’s going on?

In my case, I found that while traffic was flowing in, the click paths on the site were incredibly erratic. They’d click about here and there on the site and eventually, they’d find the store locator or contact. I look at click paths and sure enough…some of the keywords with an obvious query intent for contact, store location, etc now had click paths that were 4-5 clicks before settling in when they were previously only 1-2. Coincidentally, many of the features that these people were looking for had given up homepage real estate to some of the newer features.

The point I’m trying to make is that when time-on-site goes up, we need to be mindful of what else is going up. I might argue in many cases, longer click paths occurring simultaneously may convey the user experience for direct (or transactional) queries has been compromised. Here are some other things I’d look out for when you see time-on-site figures going up:

-Page Load Times: Are they really engaged or are they watching a load meter ooze across the screen?

-Geographic user base: Are the languages syncing up with where your users are from? Are we landing French speaking people on English pages causing confusion?

-Top Entry Pages: Are people entering on pages that don’t align with your top keywords? Perhaps they aren’t sure where/what to do?

-Most Downloaded Files: Have these changed in the time that your time-on-site went up? What does that mean?

-Internal search queries: It is funny how oft ignored these numbers can be. But when all of a sudden people are searching on changes, we may have some confused people on our hands!

There are plenty more. When I think of time-on-site, and how navigation can actually distort what time-on-site is telling us, I’m reminded of a site I used to follow in the 90’s. Go ahead, check it out (I’ll wait).


Copyright © 2008 Resolution Media, Inc. All rights reserved.