Implementing a Web Analytics Solution: Perils, Pitfalls and Practical Advice, Part 2

Implementing a Web Analytics Solution for your site can be fraught with all kinds of peril, but knowing what you’re in for goes a long way toward avoiding the biggest headaches, asPart 1 of this two-part series points out. Part 2 continues the discussion of effective ways to deal with potential problems.

For starters, there’s the challenge of campaigns.

If you generate traffic through external advertising or search engine marketing, then it will be essential to cover these campaigns in your tagging. Some vendors — HBX, for instance — are increasingly recommending “tagless” campaign conversion. Tagless campaigns are an enormous convenience — however, even they require some thought. Because campaigns can’t be setup retroactively, it will be essential for your Web measurement team to know about campaign codes before they are launched. For many organizations, that isn’t business as usual.

In addition, to get good tracking of your Search Engine Marketing campaigns — and your organic non-campaign traffic — it may be necessary to ad codes to the URLs you, or your agency, are using for Landing Pages with Google, Yahoo, etc. None of this is particularly challenging from a technical perspective, but it can take time and coordination.

Conversion Processes and Shopping Cart

In fact, no matter how your Conversion Processes are coded, they are probably going to require some extra work. With some products, such as Websidestory’s HBX, this is less true than it once was. With tagless campaigns and conversion funnels, you can often avoid any additional goal page tagging. With others, however, you are going to have to explicitly mark conversion steps and goal pages.

No matter what product you use, if you have a shopping cart on your site, then you are going to have a fair amount of custom Javascript to get the commerce component set up correctly. As is often the case with tagging, the actual Javascript required is usually quite simple — but getting your technical folks comfortable with what’s required and choosing the best implementation for your site will likely absorb weeks of planning and meeting.

In addition, internal IT organizations are notoriously uncomfortable with new technologies and will often provide exorbitant time/cost estimates for this task. This is where working with your vendor to insure that your IT group is comfortable with tagging will really pay off.

Custom Variables

Most tagging solutions allow you to set up custom variables that represent either special actions on the site, like capturing internal search results; special groupings, that is, collecting a group of links into a module for higher-level analysis; or custom interactions — with tools, for example.

The extent of custom variable tagging varies widely by implementation. For a SiteCatalyst implementation from Omniture, this is likely to be the single largest block of work. With SiteCatalyst, custom variables often capture the key analysis points — and provide most of the meaningful data to an analyst.

What can make this task especially difficult is that since the variables have to be captured in the tagging, you often have to guess at what you think will be important enough to capture in these variables without having any real data to work from.

This is a case where building in one or two iterations of tagging into your initial plan can make your life much easier. If you start with the assumption that you are going to revise your tagging after two to three months, then you’ll likely encounter much less resistance when you have to do it — as you inevitably will.

Visitor Segments

The ability to group visitors and analyze their collective behavior is one of the key capabilities of Web analytics. Most enterprise solutions these days deliver a mixture of tagged and tagless segmentation capabilities. Tagged segmentation is usually only appropriate for global segmentations that you know are absolutely critical to your business, like customer level or existing customer vs. prospect.

Tagless segmentations are extremely important for analytics — they are called ASIs in SiteCatalyst and Active Segments in HBX — and you will want to do some preplanning here to decide how many segments you are going to need and whether you will need to create tagged custom variables to help identify segment members.

There are always contract limitations on tagless segments, so it’s important to be aggressive in negotiating plenty of segments and — with HBX — ensuring that they are of sufficient duration for your purposes.

Putting It all Together

Even though Web analytics vendors have made considerable strides in developing tagless capabilities, the quality of a tagging implementation is still vitally important to the quality and capabilities of the measurement system.

Vendors have a self-interested stake in convincing prospective clients that tagging is painless and quick — and there is just enough truth in this to make their claims credible.

You can do a painless and quick tagging implementation, but you probably shouldn’t. If you have a large site and are serious about getting good measurement, you should be prepared to spend some time on all of the issues outlined above — and probably some that are unique to your site.

Investing time in this up-front process will more than pay you back when it comes time to actually producing good reporting and measurement, and will save you countless political struggles trying to re-engineer a bad tagging job.

Of course, you should always keep in mind that tagging is just the first step down a long and complex road. Managing a rollout, deciding on the best mix of tool access and reporting access, managing training appropriately, and getting your organization committed to and productively using measurement are all bigger tasks than tagging.


Gary Angel co-foundedSEMphonic, a search engine marketing tool provider and Web analytics consultancy, and is president and chief technology officer. He’s responsible for leading SEMphonic’s development of Web analytics and SEM decision making tools for Web marketing professionals. In addition, he helps companies like WebMD, Intuit, American Express and Charles Schwab maximize their Web channel marketing through intelligent use of Enterprise Web Analytics.


Read Implementing a Web Analytics Solution: Perils, Pitfalls and Practical Advice, Part 1

Leave a Comment

Please sign in to post or reply to a comment. New users create a free account.

Related Stories

Technewsworld Channels