By Andrew Edwards on April 09, 2009

Recently a customer contacted me and said her company was having a number of analytics reporting problems -- broken or missing reports, lack of trust in the data, and inability to align new reports with emerging business criteria.

"I'd like you to spend a couple of days training our staff," she told me, "and then they can take care of these things themselves."

My answer to her may come as a surprise. I told her we would not do it.

I recommended something more direct: call in an expert, not a trainer.

The most difficult challenge in web analytics today is not a lack of tool capability or a lack of data accuracy. It's how hard it is to get even the best tools aligned with a company's business goals, and to keep them aligned in a changing web environment.

If you're depending on analytics to make business decisions and you're hoping that an internal resource -- distracted with other duties, possessing not more than a passing level of knowledge about the tool and its intricacies -- can make analytics perform the way it needs to: think again. Time and again, I have seen these assumptions lead to the problems I mentioned before.

In other words, don't try this at home. I'm not going to compare analytics to a home chemistry set (it's not going to explode!), but I am going to list some common misunderstandings and how to avoid them.

Misunderstanding No. 1: "My reports aren't helpful -- the analytics tool must be broken!"

All the major analytics tools can deliver great reports these days and they can even look good to marketers. The most common reasons why the tool isn't working tend to be: lack of expertise in business requirements definition; lack of expertise in report design and configuration; and poorly implemented page tagging (flawed data collection).

Solution: Invest in expertise -- it will help you realize much more of the ROI locked inside your web analytics than you might have imagined.

An expert will be able to listen to your team's business requirements (even if you don't really know how it maps to reporting) and create a set of reports that meets those goals. If they have a technical background (and they ought to), they'll also be able to create all the tagging you'll need. They can also guide your developers through the often-tricky process of getting the right tags on the right pages or interactive modules (like Flash or Ajax). Result: clear, accurate, meaningful reporting.

Misunderstanding No. 2: "Our reporting seems stuck -- can someone just spend an hour explaining how this all works?"

Yes, someone could do that. But it wouldn't do much to help. Enterprise-level analytics deserves much more attention than an hour, or a day, or a couple of days. What's needed is a comprehensive approach.

Sure, training is an important part of it -- mainly to make sure both end-users and developers understand the possibilities and detailed tasks needed in order to accomplish the ROI. But understanding the possibilities is only the first step.

Solution: If you invest in expertise, make sure it's the right expertise. Make sure your analytics expert knows how to make it happen.

It doesn't help much to bring someone in only to talk about best practices or what a drilldown is versus a conversion funnel (and how that relates to your particular business model). It does help a great deal to have someone work with both your marketing team and your technologists, but only if it's coupled with a plan for detailed implementation, and then the implementation itself. If you're willing to invest in the right kind of expert, they will be able to build your team's understanding to a level where they can take advantage of the solution your expert ends up building on your behalf. And the better your team understands, the more likely they'll deliver ROI using analytics.

Misunderstanding No. 3: "I don't have a good idea of where I need to go with analytics. So how can I hire someone to help me achieve it?"

If you know your business, you do already know where you need analytics to take you -- you just may not know how to express it in "analytics-speak." It doesn't matter. If you work with the right expert, you will be led through a careful process of definition and implementation, and you will end up in a place pretty near where you'd hoped to get.

Solution: Trust your expert -- they know their success depends on yours.

Seek out credentialed expertise. Don't settle for someone who can't demonstrate capabilities. Look for vendor-certifications, relevant customers, references, industry recognition, then go ahead and let them do their thing. The process may at times seem mysterious to the marketer, since it can be awfully technical; or "soft" to the technician, since it delivers by extracting quantitative data to make qualitative decisions. But the right kind of expert will know how to bridge these and many other gaps. You may need to leave significant strategy, planning, and design decisions to your experts at this stage and if you've chosen them based on the above criteria, you can rest pretty-well assured they will come up with a plan that addresses your needs and is achievable. Trust them to do so.

If you're wondering where you can find qualified experts, start at the Web Analytics Association.

Analytics is complex because user behavior is complex. Your team already has a set of jobs they need to accomplish. Spreading them too thin can result in missed opportunities and lost revenue. I've often seen where companies that under-invest in expertise are the ones who complain most that analytics "doesn't really work" for them.

Don't be one of those organizations. Don't try this at home. Work with an expert in the field and measure their results.