Joe is a webmaster, and does some SEO for his own site. Some day when he woke up and logged into his Google Analytics account, he found that his site had obtained a good load of traffic, and what made him happier, he saw that the page views per visit was really high…
He started to dance.
Why not? Books out there tell you that the PV per visit metric shows visitors’ fondness for your site. The higher, the better, right?
Most of the time, it is correct to say that, but there are also exceptions, and visitor lostness is one of them.
Imagine that you come to a site with a really poor information architecture. After a couple of clicks, you totally get lost and have no idea where you are, so you have to keep clicking to see if you are luck enough to find what you need…
And at the same time, the site owner is blindly happy about his new high for page views per visit…
User lostness is a very useful concept, which was first introduced by Pauline A. Smith with Towards a practical measure of hypertext usability. It can vividly quantify the lostness of your user, and help you improve your navigation system.
The formula is very simple, as shown below:
L = sqrt[ (N/S-1)² + (R/N-1)² ]
N: number of unique web page views
S: total page views
R: optimal page views (minimum page views to complete a certain goal on your site)
The results of the formula represent degrees of lostness, which can be further explained here:
L = 0: not lost
0 < L <=0.4: no observable characteristics of being lost
0.5 < L<=1: definitive characteristics of being lost
One of the most amazing facts about this formula is – it was invented in 1-9-9-6!
Of course, no matter what a genius Mr Pauline Smith was, he would not be able to predict the whole picture of today’s Internet world. So this formula just gives you ideas on how this kind of things should be calculated, and you should not just violently integrate it into your current data analysis system, but adjust it to make it fit.
If you know what euclidean distance is, you may find the lostness formula is nothing new.It just calculates the distance between Point (N/S,R/N) and Point (1,1).
Easy but amazing, don’t you think so?