Onus Is on the Buyer to Unlock Greater Value from Web Analytics

Insurance firms must perform more due diligence if they want their Web sites to net them greater revenues. That is one of the crucial messages of a recent study from Forrester Research, a Cambridge, Mass.-based technology and market research firm, as a litany of mergers and acquisitions (M&A) and a blur of technology offerings in the Web analytics field make insurers' search for the ideal vendor a murky proposition.The Forrester report, titled "How Web Analytics Buyers Structure Contracts," indicates that too many companies are rashly choosing vendors to tackle a very important element of their success strategies: maximizing their business on the Internet.

A growing number of software companies offer Web analytics software, which provides tracking and intelligence on the number of Web site visits and visitor activities, but only 21% of companies bother to perform acceptance tests before making their first payment to the vendor, according to Bob Chatham, an analyst at Forrester and author of the study.

Further, only 28% of firms make their first payment at the contract signing rather than first demanding to see positive results from the implementation. Such behavior is risky in light of the vast amount of money firms spend on Web analytics, says Chatham, especially given that the technology's results should be taken with a grain of salt.

"There is a dawning realization that the data isn't entirely accurate," he says. "Absolute truth isn't going to come out of your Web analytics package-in terms of the number of visitors."

He says visitor counts can be anywhere from 20% to 30% off the mark. Such errors occur when, for instance, software packages recognize two different visitors to a company's site, when in fact one person visited the site from two different computers. The study's findings don't reflect well on either software vendors or their clients, notes Forrester.

Forrester points at vendors for frequently failing to include validation and auditing features in their systems. Meanwhile, close to 15% of vendors' clients never perform a single acceptance test before implementation, which can later haunt those employees responsible for the task when management starts sifting through Web analytics reports.

Chatham attributes the buy-first, test-later phenomenon to a combination of factors, including sheer laziness, blind faith and overblown respect for the vendor's name.

No silver bullet

Other reports, such as "Web Analytics: Spending, Staffing and Vendor Selection," also highlight companies' failure to optimize their Web potential.

The report, conducted by JupiterResearch, Darien, Conn., found that 70% of firms do not deliver important Web data to their companies' senior executives, while 80% fail to report data to merchandising staff.

"The biggest mistake that companies make is they expect Web analytics solutions to be a silver bullet," says Eric Peterson, a Jupiter analyst focusing on Web analytics. He explains that users believe that because the software is expensive it must surely run itself.

But autopilot, Web analytics is not.

"Companies need to be prepared to hire people to manage these applications and also be prepared to analyze the data and listen to what the data tells you," says Peterson. "The important thing is that insurance companies show a willingness to take the time to understand the application, take the time to implement the application, and take the time to use the data that the application makes available."

He argues that when a company's Web analytics software falters, nine out of 10 times the company blames the vendor, when it is actually due to the company's misguided expectations.

Indeed, vendors hardly deserve all the blame for insurance firms' woes on the Web, and many have instituted best practices, chief among them prepping the client and providing additional service.

For instance, Omniture, Orem, Utah, offers two- and four-day training courses for clients to learn the basic and sophisticated aspects of Web analytics. It also offers around-the-clock, on-demand support.

Some vendors interviewed for this article dispute the Forrester report, stating their solutions estimate online visit totals with great precision and deliver big returns on investment.

Matt Belkin, vice president of the best practices group at Omniture, indicates that one of the firm's clients made a single change to its Web site due to the vendor's Web analytics software and enjoyed revenues that paid for the software 20 times over.

And Bill Schneider, the product marketing manager of the Portland, Ore.-based provider WebTrends, says that the firm's clients have increased sales by millions of dollars due to a mere 1 to 1.5% increase in conversion (converting a Web surfer into a lead, or converting an online lead into a customer). Omniture's clients have seen similar results.

"There is so much low-hanging fruit out there it is unbelievable," says Belkin.

Who's got the cookies?

Vendors believe they have made strides in meeting industry concerns, in spite of unpredictable control variables that affect a Web analytics program.

A study by WebTrends-whose clients include Allianz Life Insurance Company of North America, Prudential Financial and Sun Life Financial-indicates that individuals visiting an insurance site block the cookies that appear 12% of the time, making it difficult for firms to get accurate Web visitor counts.

And, based on its study showing that 28% of Internet users selectively reject third-party cookies, Jupiter urges companies to exclusively use first-party cookies in data collection efforts.

Web analytics providers also are responding to insurance firms' security concerns. Insurance companies choose their software partly because they can install it in-house, which enables them to keep sensitive customer data private and in their hands, notes Jim Rose, CEO of the North Kingstown, R.I.-based Sane Solutions.

Sane counts Mass Mutual Life Insurance Co., The Hartford, Old Mutual, Progressive Insurance, Rentenanstalt, GEICO Direct, The Travelers, CNA, and Allied Insurance among its clients.

"The ability to control their own data and maintain their own security measures over this data is essential," Rose says.

"Most top-tier Web analytics solutions require their customers to subscribe to a service in which the Web analytics firm collects, stores and reports on the customer's data from a remote location. This is usually unacceptable to corporations sensitive about their customer data," Rose adds.

The vendors' efforts to cater to the needs of insurance firms have won them plaudits. Jupiter's Peterson says that the top players now have solid offerings.

"These applications all basically do the same things, and they all do the same things pretty well," he says. In fact, insurance firms have even more reason to scrutinize vendors' range of services and level of flexibility because their technology is so similar and because of the raft of recent mergers and acquisitions in the Web analytics sector.

For example, in July CheetahMail, a subsidiary of Costa Mesa, Calif.-based Experian, acquired the Los Angeles-based Web analytics company Harvest Solutions.

Chatham says when Omniture acquired the Web analytics division of New York-based DoubleClick, it extended its current rates to DoubleClick customers. He urges analytics users to include a clause in their contracts that provides an option of maintaining current pricing upon change of ownership.

Chatham recommends that insurance firms sign one- to two-year contracts with Web analytics software providers since many vendors have perished or have been acquired amid the market volatility.

Insurance companies should also work with software vendors to conduct pilots and establish a payment schedule for the Web analytics only after the software delivers demonstrable results, he adds.

Daniel Joelson is a business writer based in Washington, D.C.

For reprint and licensing requests for this article, click here.
Data and information management Claims Data security Policy adminstration Security risk Analytics
MORE FROM DIGITAL INSURANCE