Image via CrunchBase Google acquired Urchin Software Corporation, In March of 2005, a San Diego-based web analytics company. Like all web analytics Urchin Analytics had inherent errors and savvy webmasters had long known of this and made appropriate adjustments in their calculations. There are legacy issues that make it impossible to design any Web analytics that are 100% accurate. However, in the case of Google Analytics this particular inaccuracy has been purposefully left in the algorithm. The problem lies in counting visitors.

In designing the visitor count statistic Google intentionally left a glaring error from the earlier Urchin program. All international Web analytics standards count a “visitor”only when they visit more than one page on a site otherwise it is counted as a bounce.  So as an example, if someone comes to your website and looks at the first page then leaves this is considered a bounce.  If they come to the first page and go to an additional page then that is a visit.

We need to distinguish between bounces and visits because it enables us to calculate on site duration. We can calculate the time spent at a page by comparing what time they arrived at that page with the time they access the next page.  If they bounce we have no way of knowing how long they’ve been on our site. you can only calculate site duration when someone reads two or more pages.

Average site duration is an extremely important metric.  One of the most important metrics an Internet Marketer needs to understand.  Having correct data for this calculation is of utmost importance if you are going to understand your visitors site engagement.  Average site duration is calculated by totaling individual “visit durations” divided by the total number of overall visits.

The original Urchin software treated bounces as zero duration visits. Urchin used this in  calculating the average duration and the number outputted would always be under the real one.  This inaccuracy was dependent on the percentage of arrivals who bounced. Bounces are not visits and Google knows this is so, otherwise in July of 2007 it wouldn’t have changed the calculation of average duration so that it didn’t include bounces anymore.  Of course this is the right thing to do and its figures for average duration were now correct. Google put it back to the old wrong way (incorrect) one month later.  Why did Google Analytics intentionally roll this calculation back to the old wrong method when they knew it was an inaccurate number?

The senior manager of Google Analytics, Brett Crosby, explained in the Google blog that the change was undone because people complained the change meant the new numbers (correct) didn’t match with their old numbers (incorrect). This post has since been removed by Google but it’s available on many sites.

So because a few people complained, bloggers mostly, and were too weak minded and  lazy to make the appropriate changes in their spreadsheets Google Analytics  has produced an inaccurate number for “site duration” ever since.

Any metric based on the number of visits is now inaccurate if you’re using Google Analytics to make these calculations. of importance to Internet Marketers are such metrics as visit count, conversion rate and  exit rate. Google is intentionally and blatantly providing inaccurate numbers that affect all these important metrics and many more just because a few bloggers were inconvenienced by the truth.

I am absolutely aghast at the lack of understanding many Internet Marketers display when discussing their Web metrics.  I don’t know how many conversations I’ve had with other Internet Marketers where they’re bragging up their numbers from Google Analytics. They have no idea how inaccurate these numbers really are and even if they did it seems their pride is more important than reality!

The only conclusion I can draw is what amazingly low priority Google gives to Google Analytics. Can you imagine Google knowingly downgrading the accuracy of its SERP pages. Oh, that’s right, they did that in China!

In part two of this post I’ll show you how to correct the numbers you’re getting from Google Analytics. I can’t imagine planning my business around false assumptions, can you?

Tags: Analytics, Conversion rate, Google, Google Analytics, search, Search Engines, Web Analytics, Website