Monthly Archives: May 2010

Web Analytics Consulting – You Get for What You Pay

Our site analytics tool sucks – Sincerely, <Insert your name here>. Unfortunately, we have all been there. In the interest of fairness, let’s take a step back to conduct root cause analysis. Ask yourself, “Is our web site analytics tool capable of providing what we need, or is our sub-par implementation the reason why we are not able to report on critical KPIs?”

In my experience (YMMV), sub-par implementation of analytics code is primarily responsible for the majority of web analytics challenges. Even the free tool Google Analytics (designed for the “common” person) requires custom variable/parameter declarations to take full advantage of granular metrics reporting. On the other end of the spectrum, fully customizable SaaS analytics solutions like WebTrends, Adobe/Omniture SiteCatalyst, Unica, et al require extensive time to plan/deploy custom reports, visitor segmentation, visit/hit filters, and so on.

There have been numerous reports published for US web analytics forecasts, like this one from reputable Forrester Research. I’ve seen broad spectrums for web analytics projected investment CAGR from 5-25% over the next five years. The problem is, most of website analytics investment is made in the service fees as opposed to implementation consulting.

Let’s assume you sign a contract with an analytics vendor to support your flagship, enterprise sites. Chances are if you commit to X number of hours of support bundled with your MSA, you should get more favorable rates. In my opinion, you would be a fool not to pay for professional implementation consulting services. It’s the old carpenter’s adage, “Measure twice and cut once.” Complications resultant from post-production changes are more costly/difficult to rectify than those addressed in DEV environments.

Net, net all of us pay premiums for numerous insurance policies and preventive maintenance services in an attempt to mitigate catastrophic risks. Consider your investment in analytics implementation consulting services (which is a one-time opportunity cost that keeps providing incremental benefit) in the same vein as pre-paid expenses like insurance/maintenance. Worst case scenario, your analytics page tagging could cause critical failure in content delivery from your production servers for several hours (or even several days depending upon the agility of your development teams in releasing hotfixes). Please take a minute to ponder :)

Of course the previous paragraph does not apply if you’re using a log file-based site analytics tool. However, these solutions are becoming defunct faster than analog televisions. My standard retort to advocates of log file analysis tools is, “Even though you’re not prone to javascript page tags causing critical failures, you’re still subject to lost revenues from poorly implemented log file-based analytics solutions subsequently resulting in the inability to report on KPI metrics that would have manifested actionable, data-driven decisions.”

If after reading all this nonsense you still firmly believe your web site analytics tool sucks (as opposed to your implementation), please reach out to me and to the extent I am able I will do my best to help address your pain points. That’s the least I can do reciprocate your patience in getting this far. In fact, throw any interactive/online marketing questions my way. Cheers, friends!