I met the other day with a prospective e-commerce customer who is experiencing a sales conversion rate around 0.001%. On the surface, the e-commerce business is doing exceptionally well. They dominate the search engine listings including Google and Yahoo for relevant keywords and keyword phrases in way that attracts over 5,000 unique visitors per day to their site. Their website was built on a sophisticated e-commerce technology platform offering an easy to navigate and influential customer experience. From the surface, everything looked great making a low conversion rate very surprising.
Looking deeper during a quick assessment into their Google Analytics caused more confusion. Their average bounce rate was 22%, their average time on site was 11 minutes and average page view was 7. The 11 minute time created some potential concern but overall things looked good.
Suddenly it popped in my head because of other experiences I have had with clients and their website analytics that this client’s analytics could be setup incorrectly. The client didn’t have the e-commerce tracking setup and I noticed some missing code on their home page too. It started to become clear that a “garbage in – garbage out” situation was playing out. If the client’s tracking was setup incorrectly then any analysis was skewed.
Many experts instruct web businesses to regularly view their analytics and I absolutely agree. But before you start assessing you have to make sure that what you are assessing is accurate. Understand how your analytics (if you are using something other than Google Analytics) reports specific statistics. Make sure that what is being reported is accurate. For example, the site overall report on Google Analytics reports by url (the url clicked on) so if you have more than one navigation point on your home page using the same URL then the report will count clicks per url not actual navigation point. This means if you have a left rail navigation text called “widgets” with url (www.xxx.com/widget) and a widget image in the center linked to the same url (www.xxx.com/widget) then when the site overlay presents click-throughs it will count everyone associated with the URL (not he specific element) so you will not know whether the left rail navigation text or the image generated the click. Make sense?
It seems simple but when technology is involved it isn’t. If the analytics you are reading are reporting the wrong numbers, your analysis is wrong. Wrong analytics may cause the wrong strategies to be implemented to solve the wrong problem. Build a reliable reporting (website analytics) infrastructure so you see clearly into the situation and plan your next steps with accurate (or at least closer to accurate) information.