Analyzing web analytic needs
Today I was with one of our customers for a second day of WebTrends training. This particular customer manages a governmental site with the goal to inform about social topics. The training would be about using WebTrends and understanding reports. The first day we used to help them understand the tool and the basics, the second day would focus on measuring effectiveness of campaigns.
We skipped back and forward to all kind of topics, starting with explaining topic one and ending with something completely different. Tiring but it keeps my job interesting 
Deciding whether your site is doing okay or not is easy for ecommerce sites as sales and conversion are easy to measure. With content based sites it’s less easy to know whether your Visitors are feeling happy with your offering. One of the discussions we had was about Visit Duration being a metric to use as one telling whether your content or campaign is successful.
If your visitors spend an average of 10 minutes on your site, does that mean your site is optimal?
If he was on your site for 10 minutes it could mean that he is searching for the information he needs, not knowing where to find it because of -for instance- suboptimal navigation or malfunctioning internal search. On the contrary it could mean that he sees so many topics he just can’t get enough of reading all information provided.
But what if the Visit Duration would be only 10 seconds, would that be bad? With current attention span anything above 3 seconds is just great… it could mean that the information provided wasn’t what he was looking for but it could as well mean that he found the information needed and left the site as a satisfied Visitor. And what if I advertise on Google AdWords and my visitor landed at my landing page and flew away without browsing other topics. Would that be bad? NO! It most likely tells you that your landing page is perfect, the visitor has been informed.
I advised the customer to define Key Information Pages, what are the pages your organization makes writes to inform the public. Get them tagged being a KIP and see what part of the visitors saw a KIP and who didn’t. This way you get a very straight forward report; those who were informed and those who were not.
July 30th, 2007 at 4:44 pm
Using the web analytics and deriving right intelligence from it is a very tricky situation, Same thing can have two meanings as mentioned a customer spending 10 mins can be he liked the site or he spent 10 mins to look for info on the site.
The stat reports are also very important, the format , the presentation etc.
I have been using GoStats.com , they provide a comprehensive reports.