Brand Interactivism

Heroes, Discover, and Useless Vestigial Parts

It is amazing how seemingly random events can spike web traffic.

Looking at a website analytics report yesterday for Discover Magazine, I noticed that they had a huge traffic increase on Tuesday (more than double the usual.) Digging deeper, I discovered that more than 40% of that day's traffic was pointing to a single article from the June, 2004 issue: Useless Body Parts. Wondering where all that traffic was coming from, I saw that mostly it came from organic search on Yahoo and secondarily from Google.

The analytics tool enabled me to see the exact keyword phrases that people were using to find this old and archived article from Discover's distant past. It turns out that nearly all of that extra traffic was using the exact same keyword phrase - "useless vestigial parts".

Why in the world would thousand and thousands of extra visitors come to Discover Magazine's website in a single day, having searched in Yahoo and Google for "useless vestigial parts"? And then it occurred to me...

Minor spoiler alert - if you have not watched this past Monday's episode of NBC's Heroes, I'm going to reveal an extremely minor detail that should not affect your enjoyment of the show. That said, if you don't want to know anything about it whatsoever, skip ahead.

On the most recent episode, Claire (the cheerleader) is intrigued by her biology teacher's lecture about certain types of lizards that can regenerate lost body parts. They get into a discussion about evolution and whether or not human beings could evolve to do this. The teacher points out that there are many parts of the human body that are now "useless" and "vestigial".

End of spoiler - see, that wasn't so bad was it?

Given the timing, almost certainly, all this extra traffic came from interested Heroes viewers!

Still, it was strange that nearly every one of these visitors used the exact same search keyword phrase. It's impossible that just happened on its own.

Given that a majority of the extra traffic came from Yahoo and not from Google, I suspect that Yahoo posted an article on its portal homepage about Heroes with a link to its search results page for "useless vestigial parts". Discover's article shows up as the third result on that page.

On Google, Discover's article doesn't show up until the third results page. Thus, less traffic from Google.

The next step on this is to figure out how to run targeted SEO/SEM campaigns on Google Adwords or other pay-per-click search ad platforms to help capture even more of these traffic spikes. The tough part is that they are unpredictable. I believe though that if you watch your traffic closely enough, in real time, you might be able to figure out something like this with enough time to do something about it.

Perhaps Discover should be proactive about this and run search campaigns on keyword phrases related to topics that are relevant to their demographics, in particular when those topics are being discussed on very popular television shows that are watched by their demographics. Shows like Heroes.

Page Views... RIP

A few months ago I wrote about how page views were not going to be around much longer as a used metric to determine the effectiveness of a website. Well, apparently Nielsen/Net Ratings is replacing the page view metric with "time spent", which they feel is much more valuable.

Frankly, this metric (IMO) can be tough to get accurate data on. For example, if I go to an article on NYTimes.com and spend 5 minutes reading it, my time spent should be 5 minutes. But how is the website supposed to know I spent 5 minutes if my next action is to go to Gmail in my browser? I never followed up with the NYTimes.com server, so it has no way of knowing if I was there for 3 seconds or 3 hours. The only way it can know is if I return a few minutes later and click on another article. But even then it's not terribly accurate since maybe I was on a page, read it for 20 seconds, went to the bathroom, and then a few minutes later came and clicked on a new page. Was my time spent 20 seconds? 5 minutes? As far as the web server is concerned, it was 5 minutes, but that's not really true, is it?

The nice thing about page views is that it's easy to measure and objective. However, with AJAX, widgets, etc., it has become suspect as an effectiveness measurement. Time spent is a lot more subjective, but if we can find a way to feel confident in the data, is more telling about actual site usage.