RightSpot Media | Digital Performance Marketing

Five points of web analytics failure

Posted on Thursday, October 8th, 2009 at 9:09 PM by Simon Smith  |  Comments (0)

1. If you’re not measuring it, you’re probably losing money. First and foremost, we often find that companies don’t have web analytics at all or, if they do, haven’t installed systems to track essential performance indicators through their entire sales funnel. So technically, they’re using analytics tools, but practically, they’re flying blind—or at least legally blind. And regularly, once we start gathering data, we find that expensive, non-data-driven assumptions are incorrect. For example, we often find that companies spend money on search engine optimization without knowing whether their search results drive qualified traffic. Their “measurement” was simply a visual observation of high ranking. Which brings us to point of failure number two.
2. If you’re measuring the wrong things, you’re probably making the wrong decisions. Another common problem is thinking that you’re measuring effectively just because you’re measuring. In daily life, we usually know what to measure. All else being equal, for example, you’ll buy the shoes that fit. But online, things aren’t so clear, in part because there are so many marketing options, and every business and marketing campaign has different objectives. In part due to this confusion, people tend to measure the most superficial things: whether something looks pretty, has cool functionality, has high search ranking. What’s important is measuring what’s meaningful to your business. After all, pretty, cool shoes are great. But if they’re three sizes too small, you won’t walk very far.
3. If you can’t make your data meaningful, your recommendations are probably meaningless. Too often, and in large part due to the availability of free tools such as Google Analytics, we see companies address issues one and two by gathering lots of data to cover their bases. So, for example, they might create a Google Analytics account and drop the JavaScript code in their footer, thereby capturing a broad spectrum of activity. But without proper analytics installation, configuration and interpretation, this typically results in data overload rather than wisdom. It’s like installing a fire hose to get a drink of water.
4. If you’re spending more on technology than interpretation, you’re probably overwhelmed with data and underwhelmed with insight. The problem of data overload is particularly common in organizations that have over-purchased analytics technology. Carrying large analytics license fees typically means less money for analytics services and human resources. It’s a tough place to be: lots of data, but nobody to make sense of it. Analytics evangelist Avinash Kaushik recommends the 10/90 rule: for every $10 spent on analytics tools, spend $90 on “intelligent resources/analysts.”
5. If your data’s not actionable, it’s probably paralyzing. This is the other end of the spectrum from not measuring anything: measuring too much, including irrelevant details, and having no way to separate signal from noise.  Typically, this happens when organizations recognize point of failure number one, overcompensate with point of failure number four, then find themselves with a morass of data that bogs down rather than enlightens decision-making.

So there you have it. My thoroughly unscientific, entirely anecdotal and experiential analysis of common web analytics failures. Think I missed something? Let me know in the comments.

Tags: enterprise web analytics, featured, web analytics