Measure for Measure Blog

Web Analytics Environment Alert: 3 Ways to Keep Your Analytics Clean

Written by Andrew Edwards on . Posted in Digital Analytics, Web Analytics, Web Analytics Tools

wa environment

Attention Web Analysts:

Your environment is probably dirty, and you need to clean it up.

Dirty, not the way the Cuyahoga River famously caught fire back in the 1970s.

Dirty in the way that untended gardens get choked with weeds and then dead weeds and then trash and then dangerous shards of glass and eventually abandoned shopping carts and automobiles. Dirty in the way that what was once not only attractive but useful now becomes both ugly and detrimental to everyone in the neighborhood.

Here is a call to action: a sort of spring-cleaning-in-the-fall before the leaves come down. Make sure your environment is clean.

How to do this?

Let’s assume nothing about what tool you’ve got. It doesn’t matter that much. I will use language associated with Google (since it has become almost the lingua franca of analytics).

Here is a list of things to look for and fix, in order to keep your analytics from going to seed:

  1. Tagging. Long the bugaboo of even experienced analysts, both the standard and non-standard code on your pages can seem to change and become incorrect without the analysts knowledge. This is not because tags can change by themselves. What often happens is that developers decide to change a parameter and, without enough knowledge or guidance, introduce incorrect formatting or syntax into the tags. This makes them inoperable or dysfunctional. The other way tags can get “broken” is that new pages get built without any proper alerting to the analytics team and those pages don’t even get tagged at all. The result is incomplete, inaccurate reporting. The cure: a regular tagging audit of all assets.
  2. Reports. How many do you really use? Do you know how many other people are using them? And which ones answer business questions? There are plenty of standard reports available. In addition, you can create anywhere from 5 (in GA) to many more custom reports depending on your tool. Typical scenario leading to dysfunction: someone insists they need a particular report. It is built. It is looked at once and forgotten. It is never deleted. Moreover, it is mis-categorized, re-categorized and later, misunderstood. The cure: continuing review for report structure, consistency and relevance; including extinction of unused custom reports and non-inclusion of unused standard reports.
  3. Profiles. In order to look at reports, the user is provided with what is often called a “profile” which includes a set of reports and dashboards based on what should be business requirements for that user. Many users may share the same profile. The destruction of profile usefulness begins when profiles are edited by committee or by successive administrators. They become clogged with irrelevant reports, incorrect data, dead links and a general characteristic of irrelevance. They get forgotten about and then re-discovered by a hapless web analyst fooled into thinking the profile actually contains the data they were told it contained. But that was back when it was created, and now it is full of junk and nonsense. The cure: aggressively police profile production, alteration and use-cases. Remember that profiles refer to parameters that may change, rendering the profile inaccurate unless updated in tandem with the parameters.

Your environment matters. If you won’t tolerate old tires dumped in your garden, then you should not  tolerate inaccurate, incomplete tags; dead reports; or broken profiles in your web analytics.



Leave a Reply

Your email address will not be published. Required fields are marked *


HTML tags are not allowed.