Web Analytics and the March Toward Better Measurement

By Jim Sterne

The World Wide Web has generated enough tracking, testing, and merchandising tools to keep an analyst happily traipsing through data for decades. But marketers are starting to realize that there’s more here than meets the mouse. As the analytics have evolved, so has our ability to measure the effectiveness of Web programs.

The journey started in the IT department (circa 1993), where even the least marketing-savvy technician knew that a slow Web site was a bad Web site. Performance monitoring tools emerged to measure whether the server was serving quickly and set off an alarm if something went bump in the site.

When the marketing department began to understand the value that Web data could divulge about customer predilections, Web analytics was born. We have since moved from the age of demographics, through the age of segmentation, and into the age of behavioral analysis to gauge how individuals respond to different stimuli (advertising, marketing, and customer service).

More than a hash mark on the doorframe to indicate growth, this new generation of Web analytics, full of rich, granular, and instantaneous data, now offers a window into the hearts and minds of the marketplace. It also improves our ability to capture those hearts and minds. We’ve come a long way — and we still have a long way to go.

We now can collect more data about an individual’s activities on a Web site than can possibly be put to use. We track every page viewed … every element moused over … every menu dropped down … every key pressed … every video watched. Browser-based, client-side tracking techniques can collect the most complete picture of what a visitor does on a site — even capturing data entered into a form that the visitor chooses not to submit. On-site search analysis reveals what people are looking for that our menus and links and buttons have left un-findable.

Search marketing has evolved from measuring sticky eyeballs to gauging profitability. The whole reason Google is giving away its Web analytics service for free is to prove that the clicks you pay for result in sales. As a result, cost per click is being replaced as an online marketing metric with the more relevant cost per sale.

 

It’s the Behavior, Stupid
Behavioral analysis tools, which give us insight into the intent of the site visitor, are another blade for slicing, dicing, and measuring performance. It’s great to know that a single campaign or a specific persuasion scenario resulted in a 10% lift in sales, but when you can determine that the most profitable buyers were the ones who viewed specific Web site content, you can more finely tune your promotions to attract just those types of prospects. With a database of these types of behaviors in place, we can take steps in real time to influence an individual’s activities to increase sales, prevent attrition, or build loyalty.

We are moving from one-size-fits-all to literally putting the right message in front of the right person at the right time. There’s an 85% probability that Nancy will purchase this product? Let’s dynamically generate a free shipping offer to seal the deal — and push up our conversion rates. This is the type of emerging analysis that leads to strategic insight and competitive advantage. (Stay away from dynamic pricing, though. The transparency of the Internet has largely taken away our ability to play with pricing across markets. Amazon toyed with dynamic pricing based on behavior and purchase history. Within days, it was publicly outed and Amazon wisely backed away from this tactic.)

An integrated tracking system can chronicle what attracts the most profitable Web site visitors, how they wander through the site’s pages, and how they do or do not buy. This system can feed the results back to an advertising system so that the most profitable promotions are automatically repeated. It can send results to the Web site design team for insights on how to improve navigation. In other words, it can have a significant impact on sales and profitability.

As Web analytics grow more sophisticated, so do marketers’ ability to correlate online and offline activities. A postcard can drive traffic to a specific Web site page, verifying its effectiveness. An online offer can result in a printed coupon that’s only good in the physical store. Through these activities, the onetime fantasy of total customer behavior analysis is a half-step closer to reality.

Companies are experiencing dramatic sales lift when their messaging is aligned across all media, as customers don’t think in terms of channels, only brands. Monitoring the offers that yield the best results in specific media combinations and sequences points a very clear finger at continuous profitability improvement. Firms that track customer satisfaction scores — and correlate those findings into their action plan — can enjoy an additional boost to the bottom line.

These concepts are not far-fetched and they’re not far off. The tools and techniques are available now. If you are one of the pioneers who has figured out how to pull them all together and (here’s the tricky part) are willing to share your experiences, I’d very much like to hear from you.

Click here for a glossary of web analytics terms.

 

Jim Sterne is an Internet marketing consultant and president of the Web Analytics Association. He can be reached at jsterne@targeting.com.