Posted by Andrew Edwards on March 7th, 2011 at 5:04 pm

It seems nearly every confab of analytics folks (or digital marketing folks) features at least one and sometimes more than one "marketing book" on offer as a giveaway or a "show special".

Having seen many and read some of them, I have noticed a pattern: many (but not all) spend a long time telling you how what they're going to tell you is going to change your professional life. Then they tell you it's all about "you" and how you'll go forward newly confident having read the book. Somewhere in many of the books--and I do not tarnish them all, especially since Jim Sterne's books are substantive--there may in fact be a few pages of actual specific advice; then the rest of the book repeats that advice over and over and pumps you up with suggestions about how astonishingly different everything will be if you just would follow those two pages of advice stuck between page 254 and 257.

The reason I mention this is because I will now discuss a method that (in my totally biased opinion) could be the heart of a good, long, expensive book about how to achieve web analytics success. Except that I would rather just make it a blog post.

Here it is, the heart of the book I chose not to write at this time:

The 5 Steps to Web Optimization Success

e5o: eBusiness 5-Step Optimization

e5o is a manageable, repeatable process for web analytics/web optimization success. I have seen it used to great benefit and ignored at some peril.

While not complex in structure, it is more difficult to carry out in practice. That's because web analytics success requires the participation of more than one business unit. And it may also require the help of experts not found inside the organization. Nothing I say here should be construed to imply that "it's easy if you just do what I say".

In short, e5o describes what I believe are web analytics best practices based on thousands of engagements over the course of nearly ten years in the field. My company, Technology Leaders, uses this method as often as we can convince our clients to embrace it. When they do, the process is much more clear and effective than otherwise.

The e5o strategy can be deployed in situations ranging from traditional web analytics through user testing and into social media tracking. e5o enables analytics customers to define steps and evaluate the success of any measurement program no matter its goal.

Very simply then, here are the 5 Steps to Web Optimization:

1. Define Drivers (KPIs) - Hold facilitated discussions with key stakeholders to align business goals with measurement architecture

2. Build Metrics (Deploy Tools) - Analytics experts (in-company or outsourced) work with analytics tools (WebTrends, Omniture, Google Analytics and others)--by devising report design, tag specs and close support for developers

3. Plan Actions (Analyze & Recommend) - A properly configured analytics tool now delivers the reports needed by the customer and constituents in order to facilitate action-definition--in other words: "it says the page withe the orange button is tanking--should we try some different button placements?"

4. Create Changes (w/Agency/Developers) - The technology team hands off the findings and recommendations to the creative groups at the customer and its agencies with an eye towards measureability of changes. Presumably they come up with a wave of new creative to be tested for effectiveness.

5. Measure Success (Iterative Optimization) - You've already got the measurement platform--so now measure the new creative. Did the fresh creative help, did it hurt? Keep refining the "better"; throw out the "worse". Fancy words for this include "A/B testing" and "multivariate testing"; it can be automated or not. But this is the part that continues to "live" until you've gotten the best possible results out of your current site configuration.

For any new initiative, use e50 again: wash, rinse, repeat.

I may have raised more questions than I have answered, and I plead guilty if that is the charge (were you expecting a book?).

Seriously--the abovementioned 5 steps really work when understood and properly staffed with a cross-disciplinary team of internal and external analytics experts and creatives.

Putting together that team? Important, but not the subject of this post.