1. Timer Ticks

So you want to know how much time a visitor spent on your site, or the average amount of time spent by a segment of visitors.  OK, not much of a problem.  But now you want to know how much time was spent on a page.  Still not hard IF your page is a page.  If your site is in Flash however, you have a stickier issue.

Now you already know how much I love Flash (sarcasm), and here's yet another reason I'm not a fan of Flash sites that aren't coded properly.  A Timer Tick basically counts the elapsed time that occurs while viewing a Flash "page".  You can set it up to increment for whatever interval you want (say 10 seconds).  The problem is this "increment" isn't captured in a Time on Page metric.  It's captured as a click.

Because of this frequency of clicks, all metrics around this page are skewed.  If you look at a link report, you will probably say "Whoa - that site/page had a lot of activity!"  The problem is that no real links were selected in the visit to this page.  Instead you have clicks, lots and lots of clicks; all from the timer tick.

Sure, using clicks to track activity can be a handy trick in a pinch, and might even be recommended if there are a minimal number of links on the page.  But as an analyst who relies heavily on the accuracy of my data, I hate anything that artificially skews a report - anything that pretends to be something it's not, which brings me to...

2. Fake Pages

Another pet peeve of mine? Creating fake pages that only redirect to another page - for the sake of tracking clicks.  Instead of a page masquerading as a click (as in the Timer Tick example), you have a visit masquerading as an incremental page view.

Here's what usually happens, and how you avoid it.

  1. You have a Friendly URL you want to track
  2. You direct that URL to a page on your site
  3. That page redirects to the landing page you want the visitor to go to
  4. You count the views of your fake page to know how many visits came from your Friendly URL (say it was  print ad).

Why is this bad? Let me count the ways:

  1. It creates artificial page views (which skews your metrics in this regard, analysts hate that)
  2. It doesn't count those visits under a source you can classify (usually falling under "internal referrals" or "direct traffic"
  3. It is completely unnecessary

With the right process in setting up friendly URLs you can measure those visits specifically against the campaign you created, you don't have to set up a fake page and skew data.  It's simply a matter of adding the right tracking code to the URL string that the Friendly URL redirects to (in step 2).  That's it.  Most web analytics tools can do it.  And Google does it best.  But that's a topic for another post.