Thanks for your comments.

In this case the source of the data was from Nielsen Online, so it’s based on consumer panels and not different web analytics tools.

While panels are not perfect, the results are directly comparable. So what we can say with confidence is that these sites are top converting websites. Whether the percentage numbers are correct is somewhat academic: this is about ranking.

Rankling the top sites using the same methodology against other sites is inherently useful for marketers simply because they want to know which the top converting websites are. We all look to learn from what other sites do well, so it’s natural to want to know which the top converting sites are.

This research was primarily about the techniques used by the top converting sites, highlighting examples that online marketers can learn from. Get the free ebook and you’ll see this.

In a previous blog I have cautioned against using the conversion rate for benchmarking . It fluctuates wildly based on seasonal factors, promotions and a host of other factors. For this reason the most valuable way to think about a websites conversion rate, is to look at it over time, comparing it to part performance in the context of promotions, seasons, launches and changes to the website itself.

Don’t get me wrong, I’ve been in love with Conversion Rate for 12 years, but like many marketers I’ve started to realise it has severe limitations.

Without knowing what their average order value is, what it cost them to get an actual customer/sale, how they actually measure conversion rate and even what type of analytics package they use, such comparisons are really meaningless.

We all know that different analytics packages all have their own different ways of counting visitors, unique visitors and absolute unique visitors, so unless they all use the same package, the figures are already flawed.

And, depending on what your selling, average order values could range from $5 to $5,000. And on that basis a site with a 2% conversion rate could still be 10 or even a hundred times more effective than one with a 20% conversion rate.

Plus are they measuring sales they got offline that were generated by the site, plus sales online that were generated by offline means?

There’s just far too many variables for us to keep using conversion rate as a means of comparing our sites against other sites.

By all means use it to measure your own sites performance and try and improve it, but to use it as any form of outside comparison is ultimately useless.