I am just finishing up a case study for the next issue of BI Review on FoxSports.com and the ways in which they are using Web analytics to make their content more relevant to sports fans. Given the priority businesses of all kinds have placed on making the Web the primary customer interface, I think it's a very compelling story that's even a little exciting - and I don't get much excited these days. So, you can call this column a plug, but I think many of the things companies like FoxSports are doing with Web analytics will trickle down to your point of view sooner or later.Going forward, Web sites won't be the subject of periodic re-launches; they will be dynamic documents that change as quickly as information dictates. Ideally, the Web interface should be designed by the customer, not at a table of people sussing out wire frames and arguing over font sizes. Web analytics are the ideal mechanism to make this happen since they monitor and analyze a multitude of behaviors and interdependencies. Yet most Web marketing analysis has been centered on advertising, subscriber signing and incremental revenue building. Trust me, consumers are not patient or forgiving when it comes to his or her surfing habits, so you'd better start listening to them.
Bill Gassman, Research Director at Gartner Inc., looks at the Web analytics market as a sort of two-humped camel. "The right hump is the people who get it, they can take information and run with it. Real time actually matters to them because they can make and implement a decision and process does not get in the way, it helps." As Gassman will confirm, Web analytic tools have matured to the point where the information is not necessarily 100 percent accurate but it's relatively accurate and you can do a lot of things with it.
Those that do that are making some major progress, but then there's the left hump of Gassman's camel, where people are looking for "comfort information" on the Web. "They are really struggling to even understand what's going on and when they have some insight they have problems implementing things just because their processes are poor. They just spent three months getting the Web site perfect and now they're going to let it sit for six months." The left hump is definitely speaking for most businesses today.
Gassman's observations on process are spot on. Web analytics in the hands of an un-empowered analyst are no more useful than predictive analytics are in the hands of a similarly powerless business analyst. The effects of what I am starting to call "culture freeze" are even more noticeable on the Web since the control of Web design usually rests in the hands of a small group that at best can try to interpret the goals of the business. In the case of the Web, things move much too fast to work successfully this way, and as companies move away from old contact models, they need to apply resources to be effective from the process standpoint.
Some of this can be automated with business rules. Many of the higher-volume media sites automatically rank stories based on popularity or how often the story was emailed. Better yet, they make the same kind of information available to the consumer in lists that can be a key driver of traffic and extended stays onsite. Time spent on site is one of many key metrics employed on FoxSports.com. Yet media buyers prefer to look at total impressions, not where visitors came from or went to or how the materials on a site led a visitor from point to point. At FoxSports, analytics are part of the editorial process and cycles that are not only monitored minute by minute, but contribute to new product development and even help manage the staffing mix. The same and more measurements are being applied to new rich-media for audio and video, and for Ajax-based updates for box scores and play-by-play.
You don't have to be a media company to profit from this thinking. How many businesses do you expect actually react to external events when it comes to their own Web site? How quickly do software companies reposition products after a security scare? How many battery distributors update availability by location during the hurricane season? The answer appears to be some, but certainly not most. The typical response is a press release, which is useful to a PR firm but not so much to a customer with wallet in hand.The same kind of reflexive response management wants on the supply side is greatly abetted on the demand side by Web analytics, yet outside media and some other short attention span organizations, the demand side is underserved.
The demand side is also a mostly passive participant on the Web. Megan Burns, senior analyst at Forrester Research, likes to stress not just behavioral, but attitudinal measurement on the Web. Where you do see customer opinions, surveys and satisfaction rankings in Web analytics, they are usually used to address an individual's performance, not to develop new or better products. But when customers are fending for themselves, behavioral analytics are valuable on their own. "It's a safe bet," Burns says. "If you successively refine a search five times in a row and found 40 percent of the time the visitor got zero results, chances are good your site isn't using terms that resonate with the customer or you're missing significant content they are looking for."
In relative ignorance, I had been wondering why Web data had not been better incorporated as a feed to data warehouses. As it turns out, Web analytics are mostly provided on demand using hosted data and software as a service, since each page view can generate three or even 10 times the amount of data as is collected by transaction systems in credit card and other industries. This creates huge problems for traditional OLAP and the DBAs who manage them. Beyond the volume of event data, there is high cardinality, meaning the lists in the data tend to be very long. I talked to Jim MacIntyre, CEO of Web analytics specialist VisualSciences on background for the FoxSports story. I found his observations so interesting that I'll be interviewing him in the September issue of DM Review.
The bottom line is, if you deal with the public or have a large internal organization to manage, you need to learn more about Web analytics. The pursuit of this, by the way, might be a good career path for that shiftless nephew of yours, since a central obstacle to leveraging the technology is a dire shortage of good Web analysts. Gassman finds the best candidates are loyal internal folks, who often work in a marketing function. "A good analyst can not only build a spreadsheet, he or she will tell you why it's there and what you can do with it. They're cheerleaders to help marketing people use the products." Judging from what I have learned in the last two months, you can do a lot more with Web analytics than just email a report on ad fulfillment.
Watch for the FoxSports story in the September issue of BI Review. If you don't get our magazine already, be sure to subscribe at www.BIReview.com and feel free to contact me with suggestions and feedback at firstname.lastname@example.org.
Register or login for access to this item and much more
All Information Management content is archived after seven days.
Community members receive:
- All recent and archived articles
- Conference offers and updates
- A full menu of enewsletter options
- Web seminars, white papers, ebooks
Already have an account? Log In
Don't have an account? Register for Free Unlimited Access