Myths of Web Analytics    1    2    3

Myths of Web Analytics:
The Myth of the Average Visitor

By John Marshall

Once upon a time, in a faraway land called dotcom, there was a web marketing maven whose job it was to analyze her company's web site. She loved using the tools provided by the IT gnomes--she just knew that all those pie charts had to mean something, if only she could work out what. Like all others in dotcom land, she loved to see the little slices of pie get bigger and smaller, all the while basking in their glorious 3-D color.

She was dazzled and even delighted, but somehow the truth behind the data always seemed just out of reach. Her problem, dear reader, is that she had read The Visual Display of Quantitative Data and she knew the dazzle was not the truth.

The marketing maven decided to seek the advice of the wizards in the ClickTracks castle. They also knew a thing or two about data, and sternly warned her, “ClickTracks say: Beware of averages, or they will get you."  While this made the marketing maven a tad nervous, she asked the wizards to explain.  And this is what they said:

A Single Number is Not the Answer

"You see, businesspeople are always under time constraints—they want ‘one number' that'll drive success. Analytics software makes it easy to produce a nice average because it can perform enormous numbers of calculations quickly. Unfortunately, web analytics data is highly variant and a single average tells you very little."

"Instead, you've got to focus your decisions around the averages for two or more visitor groups. For example, the Average Time On Site tells you almost nothing when used alone. It just isn't that interesting that organic search visitors spent an average of 54 seconds on a landing page. When compared across PPC vs SEO visitors (using labeling of course) it reveals a great deal about those two groups.  It IS interesting that organic search visitors spent 54 seconds on a landing page when you know that PPC visitors spent, on average, 123 seconds on the same page.  Might want to see why there's such a big difference?"

"Keep in mind that ‘the average visitor' doesn't exist. You instead have different groups of visitors that are constantly in flux--this means that they may have an attribute such as Average Number of Pages Viewed that is higher for some than others. The best data comes from such comparisons, and from frequently exploring the visitor groups by defining new types of labels."

"Now I understand," said the marketing maven.  "It's more important to understand how groups of visitors compare to each other than it is to try and lump them all together!"  Hearing this, the wizards took a moment to confer; then one spoke: " That is enough for now. Our work is done with you, young marketer.  Go forth and better understand your visitors by separating them into groups."

And the marketing maven did just that!

continued >>

Myths of Web Analytics    1    2    3