I’m working under the assumption that most of the people seeking a definition these days are the ones who want to understand the value of analytics, or the output/product. I worry that the definitions that are out there today don’t offer the world any real explanation of what they can expect to receive, and often lead businesses into making poor hiring decisions for these positions: people who are overly engaged in the statistics, models, research, etc., but who don’t “look up” often enough to see the world around them. Yes, these are requisite skills, but they don’t produce a usable output if the analysts aren’t in touch with reality (what’s happening in the real world, what’s being talked about at investor meetings, in boardrooms, and in the news, which is rarely web metrics).

In that sense, I do think that these definitions are capable of doing harm, but maybe I’m just being dramatic

I wanted to clarify that I’m not saying either of you are arrogant or self-serving. I know you both, and know that to not be true.

Rudi:

With your definition (and almost everything else on webanalyticsdemystified.com), I would just say, “to what end? So what?” I think the definition is fine, but doesn’t capture why the work is being done or what the value is. It’s a recipe for work, with no opportunity to taste.

Stephane:

I agree that your definition captures the scope of skills that it takes to get this work done, and does address the, “So what?” question, but I would say this is a great definition for our own audience, perhaps less for an executive audience. Also, I think in the outcomes/purpose, I’d add that there are many implications beyond ROI and online activities.

My point is that the definitions we pose are arrogant and self-serving, and far too tactical to mean anything to the greater world out there. We aren’t defining web analytics for ourselves: we know what it is. A definition needs to convey meaning to people who are unfamiliar with (or misunderstand) either the purpose or potential of web analytics.

While mine may be simpler, it is designed to mean more to any CEO, CMO, or CFO out there than one that tries to stuff all of our intellect and awesomeness into a bottle.

A pilot can tell you that he flies a plane, or that he works rudders, throttle, routes, GPS systems, fuel systems, a radio, etc. etc. etc. All the world cares about is the plane flying part.

Strangely enough, although the simple definition seems basic, it has a far wider scope and conveys far more value. Specificity and detail just puts property lines around the scope of our work, when the truth is we bend and stretch to nearly anything.

I touched again on the subject last August at http://blog.immeria.net/2009/08/definition-of-web-analytics.html and came up with the following definition:

“The extensive use of qualitative and quantitative data (primarily, but not limited to online data), statistical analysis, exploratory (multivariate testing) and predictive models (behavioral targeting), business process analysis and fact-based management to drive a continuous improvement of online activities and improved ROI.”

The whole explanation and logic is posted in the two posts mentioned above.

I feel your definition is interesting but might still be too simple to convey the scope and complexity of true online analytics.

Regards,
Stéphane

Interesting read, although I am not sure of the definition.

How about:

The process of measuring, analyzing and interpreting the results of interactions between users and websites in the context of specific goals and objectives.

Swiped partially from this post!

http://john.webanalyticsdemystified.com/2010/02/10/defining-social-marketing-analytics/