For those who use Blogger as their delivery platform, a few months ago they added a Stats tab so that people could see how many visitors and how many articles have been read and when. Because of this, I decided to add a new widget to my blog showing the Top Ten posts based on readership. As an experiment, I've been tracking the top five and seeing their movement.
Not surprising, some of my more recent entries have been performing well, but the top spot has been solidly for the past few weeks an article I wrote about pairing with a domain expert. I thought, wow, that must have been a pretty good entry (it has the benefit of being somewhat brief (LOL!) ), but then, as I was looking at the distribution of the stats, another thought came to me... is the fact that this post is in the number one spot because it's the best article on my site, or is the fact that it's being shown as the number one article driving its continued stay at the top?
I've only been keeping track of this stuff for a short time; I don't have a large volume of data to work with, so this may all be totally moot in a week or two. For now, it seems to be holding true. Also, it's entirely possible that there may be other aspects driving it to the top position, such as re-tweets, other blog posts mentioning it, etc. The key point though is that the traffic for the article increased after I put up the widget. It reminds me of when I was a kid and I'd get to see a copy of Rolling Stone and see what albums were in the Top 50 spots. Was I influenced to buy the number one album because it was the number one album? Sometimes yes. So in this case, an external indicator drew my attention and cause me to think "hey, maybe I should check this out."
So what does this have to do with testing? My point is that sometimes all it takes is a little change to an order, or a new way to display something, and people will respond in a different way and use the application differently just based on that one change. Our plans and our expectations need to change accordingly; just because we think we know the layout of our application and how it's used, just some simple under the covers changes can alter the playing field and bring to the fore things we might not consider. Likewise, just changing the order of a display can bring new activity to areas that may not have been considered or had little if any attention paid to them.