Are we still lying with statistics in the internet age?

In the course of a search engine optimization project, I’m often asked during the competitor analysis phase why one website ranks highly in Alexa, while another ranks highly in an other statistics supplier’s top websites survey.

As our two part article Web statistics for internet market research: pick a number, any number illustrates, there are a plethora of services offering web statistics. Website owners often cite their Alexa rank to demonstrate how much more important their site is compared to a competitor’s. Other website owners will pay for Nielsen//Netratings panel research, a sampling technique born in the 1930s which surveys about 0% of Italy’s adult population – no wonder then the IAB has called this technique outdated.

Most of us don’t have a background in statistics. We easily fall for the authority of pretty pictures and lines that go up and down. Darell Huff covered the topic well in his 1954 classic, How to Lie with Statistics [UK edition]. That his book is still in print is a testament to the power of his message. More recently, Edward Tufte has tried to teach us the errors of our ways in his renowned The Visual Display of Quantitative Information. [UK Edition]

The problem with most of the available Internet statistics is that the supplier’s methodologies often lack scientific rigor, a fact either lost on or ignored by data consumers. In other cases, data collection and processing methods are inadequately documented to allow for external scrutiny. Hiding behind declarations of “proprietary techniques” is no excuse – the underlying techniques of statistics sampling and elaboration are well known, as are internet data collection methods.

Unreliable web statistics lead to unreliable internet business decisions. Before your company uses marketing data on Internet usage in your business processes and planning, consider what limitations might exist in the data, and if these are limitations you can live with.

It is our hope that industry associations like the IAB and the Web Analytics Association (WAA) will spotlight the need for modern methods and accountability in internet statistics gathering and reporting.

See our full article on web statistics for a more in-depth discussion of this topic.

Similar Posts:

• Web statistics for internet market research: pick a number, any number
• Eying Search Engine Market Share in the era of Bing
• Google Trends for Websites, now with less data
• The new email marketing metrics
• Google Labs has re-released the much-maligned Google Web Accelerator

Registration is now open for the next SEO Course (March 22 and 23) and Google Analytics Course (March 14 and 15) in Milan. Don’t miss the opportunity!