And…you may find yourself surrounded by LCD screens loaded up with graphs, charts, and tables. From time to time, your boss or another person in the office might ask you for a snapshot of a page, or maybe wants to know the list of most popular products which should be considered for the top real-estate on your Home Page. Could be that this is all they need. Chances are, if this is your day, and according to some reports this is as high as 65% of our active practitioners, you may not yet know the power in process adoption.

If you work in any one of thousands of small businesses in the world making an investment in web analytics you may be the only means to make that investment work. Let’s face it, many times, companies adopt technology or methods for poor reasons. I’ve actually been in board rooms and city manager’s offices where questions of accountability were raised and highly educated, seemingly independent thinkers explained their decisions by how another agency had engaged in similar activity. Worse still, is the fact that when pressed further, it was also offered that they had not been given any data to support. In FEMA post-disaster remediation efforts, decisions like these previously cost cities in South Florida millions. For a small business, each decision could mean the difference between stretching to meet payroll or blasting out walls for expansion. Here I am giving some really simply defined means to taking process analytics into your own hands.

Warning: What I am about to discuss may increase your personal business contribution value and present you with opportunity. Be prepared…

• Success as a web analyst is inevitably tied to the ability to produce analysis which yields insight to drive business decisions as they relate to site/business performance metrics. Each metric needs to be tied to a goal. There is no room for arbitrary measurement. Boil all the numbers out of a particular function of your website. Know who your visitors are, where they come from, why they visit, and what makes them act. Place this in the context of each of the purposes for your sites existence and get an idea of your relative success.
• Reports should describe, both visually and through written content, exactly how each measurement is impacted by forces. Each major facet of the business should have a sub-report. Each sub-report should have a graphic; at the very least a table. These should put the measurement into context. The written analysis for each section should have a few key observations and qualifying statements. Each paragraph thesis should be followed by support and suggested actions. When possible, isolate the potential value of each change.
• Disseminate reports to each major party with an interest in the metrics described. At first outing, give everyone with a stake in the metrics an opportunity at having your information and analysis at hand when calling the shots. It helps identify key people geared toward helping you achieve wide adoption. Then, start putting the reports in the hands of the people whom you KNOW need your data.
• Take ownership of the numbers, delegate co-ownership to key parties. Don’t wait until your third consecutive report on marketing efforts failing to start to shake up that department with a little rhetoric. Its business, and the livelihood of dozens, maybe hundreds of people. Mention, specifically if necessary, that certain actions are costing a great deal with no return, or that the conversion value of an adopted practice is not worth continuing.

• Theorize, support, examine, & experiment. A big part, I understate…an ENORMOUS part of being able to effect change is to get a sense of how the business machine is affected by actions. Start pairing actions with numbers. Start mapping statements like: ‘When x is y, z occurs n%’. Get an idea of how to impact x and y to ultimately increase the percentage of z. Build a solid understanding of the dynamics of the relationships between the numbers. What can you do to influence visitor behaviors to improve value? Tests can help to expose these behaviors as:
  • Existing - Is this something isolated? Is it phenomena? Is it human?
  • Relevant - Does this matter? Why? Where else is it? What does it tell us?
  • Conditional - Can it be influenced? Can I build a lever for this?
  • Valuable - Will endorsing this return value which will justify resource costs?

• Collect Feedback on the Process, Follow-Up, Apply Method to Resource Allocation. A process is only as good as the communication between its components. Its AS important to collecting data and creating experiments to continuously purge a chain of command of any noise. Things will happen. Internal politics, office anxiety, creative differences exist in every workplace. Every task passed down from the report is going to eventually be met with a person with a history and an opinion. The power and grace of an analyst is their adherence to data and the ability to present this in such a way as to connote a form of passive leadership. So long as measurement was conducted responsibly and the logical connections are in place, there should be few chances to become personally involved. This won’t stop people from trying. Send the data out, check up on the issue. Then, when things get gnarly, invite a ‘mediator’ to participate in a session to gather perspective and translate semantic differences.

If you follow these means, your process will have moved from ad hoc reporting and paper-wasting to becoming a valuable cog in the business decision-model. These might be sort of incomplete, but I think they outline some pretty important universal components of a solid ‘implement it today’ type of process. Used wisely, these should help immediately infuse the thought-path necessary to arrive at a more agency-specific process. For the most part, I would say these are loosely related to some of the major components which were described by Eric. I’ve used a little license as they’ve applied to my experiences.

The bottom line, is, that it doesn’t take a Fortune 500 budget to achieve this. In fact, it can be applicable to businesses as small as 4 or 5 people. Its a system. Its scalable, dynamic, and self-nourishing. It requires a human for interpretation. Resign yourself to this mantra:  Be the process, Be the process, be the process…..