Getting in Touch with Your Inner Voyeur

FRAMINGHAM (08/02/2000) - A small town in anywhere, USA, a man eats his breakfast every day at Lulu's Diner. He spends his US$3.95 there--and not at the restaurant across the street or the new chain out on the highway--because the waitress knows he likes one of the yolks broken but not the other, butter not margarine, his coffee black. Why go elsewhere?

The simple premise that keeps that customer at Lulu's--get to know what your customers want, then give it to them--has been edging closer and closer to technology. And the result--personalization--is something analysts and vendors from here to eternity have been quick to label the next big thing. But personalization, like microbrews, comes in many varieties, including greeting customers by name as they sign on to a site, targeting e-mails with special promotions and altering homepages so that different users see different displays. And if you peel away the hype and step back from the flashy technology, a couple of questions remain: Does every customer really want to be a Norm in his own private Cheers? And even if the answer is yes, is personalization technology really worth your while?

Maybe, but just as likely, maybe not. So the challenge for companies that have not yet joined the personalization throng is this: to draw a careful line between serving the customer and offering services the customer doesn't really want or that won't make the customer more loyal. And then to fall on the right side of the line.

YOU DON'T KNOW ME, BUSTER

Personalization technology--which profiles customers either with their input and consent or else inferentially (often invisibly), through their behavior or buying histories--offers a more precise way to match products or services to likely buyers. Its rise to prominence occurs now for a number of reasons.

First, the advent of the Web has led customers to expect a level of service from e-commerce at least as high as they would get by dialing a call center or shopping through a catalog. "The possibility of that dialogue raises the expectation among customers," says Frank Ingari, founder and chairman of Wheelhouse Corp., an e-services provider in Burlington, Mass.

"The art of the possible has been transformed," agrees Marc Singer, a principal at McKinsey & Co. in Redwood City, Calif., who notes that the Web offers the customer the opportunity to engage in the personalization experience--this is who I am, this is what I like--rather than sit passively on the receiving end.

But even as the Web expands possibilities, it limits them, and that presents real challenges to companies wrestling to find the best ways to reach their customers over the Web. "Shopping in the real world is a rich, 3-D experience," says Steve Larsen, senior vice president of marketing and business development at Net Perceptions, which makes collaborative filtering software. "When you reduce that to a 14-inch screen, you eliminate a lot of that experience."

Businesses are now required, rather than merely encouraged, to make up for the restrictions of a new medium.

Add to those factors the fleeting loyalty among Web customers--"It's more difficult to gain trust online and far easier to lose it," says Larsen--and what emerges is a rich-soil ideal for personalization technology, with plenty of hype and money to fertilize it. A March 2000 report from Forrester Research in Cambridge, Mass., says companies spent an average of $24 million on e-commerce infrastructure in 1999 and will spend $41 million in 2000. While it's hard to separate exactly how much of that personalization accounts for, its growing popularity suggests it's a sizable chunk. And in the same report, personalization ranked third on the list of the most important software purchases for the companies surveyed.

SUPPOSE YOU'RE NOT A BOOKSTORE?

For obvious reasons, personalization technology has sown its seeds most widely in the retail industry, where shoppers move around often and leave behind a trail of browsing-and-buying crumbs that marketers can collect and analyze.

Most people who have lived on this planet for more than a minute know about Amazon.com's book recommendation service, which uses collaborative-filtering technology to match customer profiles with those of other like-minded customers and recommend books accordingly. In a brick-and-mortar world ruled by superstores, Amazon.com has found a way to do what the nearly extinct local bookseller used to do--get to know customers and then help them get to know books.

Does personalization make sense for every business? Probably not. But the fairly narrow set of uses to which personalization has been put so far is certain to expand. In the meantime, you shouldn't sign any big checks for the necessary technology without first asking some key questions.

Do I know what my customers want? The first step in assessing personalization is to take a close look at the information needs and privacy concerns of your own industry's customer base. Sensitive health-care data is just that--sensitive. Few people want their Viagra prescriptions posted on the bulletin board down at Lulu's. But an online pharmacy that lets Grandma's caregiver know when her blood pressure medication is running low is performing a valuable service. People are similarly guarded about financial data, but if the information an online bank has about them means they can view accounts online in a secure environment and receive offers for relevant services, so much the better.

Do I know what I want? McKinsey's Singer figures that around 90 percent of visitors have left a website by the time Page 3 comes along. If your goal is to convert as many as possible of those 90 percent into first-time buyers, you need to concentrate on designing the content of those first few screens by figuring out who is visiting your site. A bank with an online presence, for example, will want to lavish less attention on a college student with a beer-money savings account than on a thirtysomething who might be shopping for mutual funds or a home mortgage. Personalization technology could allow a site to segment those potential mutual-fund buyers and carry them--invisibly--through a different buying experience.

Can I integrate? "Personalization needs to be part of an overall channel-management strategy," says Randy Covill, senior analyst for e-commerce applications and strategy at AMR Research in Boston. "It's a wonderful opportunity, but only if you can relate it to the rest of your channels." And there's the rub. Pure dotcoms, starting out with a fresh customer database and brand-spanking-new information, might have an easier time. Established companies with multiple channels--call centers, brick-and-mortar stores, mail-order catalogs--could alienate customers by keeping them in unlinked silos. What if a music fan orders six CDs over the Web and a week later receives a phone call on a Saturday afternoon from the same company wondering why he hasn't ordered from the catalog in months? Besides annoying the customer, the company has just flushed call center resources down the toilet.

Can I afford it? There is an ever-growing collection of vendors ready to sell you their personalization wares, some of which come with hefty price tags. But you may have what you need already; the longer you've been collecting customer data in any form, the more ready you may be to turn it into an overall personalization strategy--as long as you have the necessary expertise in data warehousing and data mining to help you analyze what you've been collecting. If not, it's worth looking into application service providers (ASPs) that will collect and handle your customer data on their servers. And a new crop of data brokers, which collect customer information from outside sources and then sell it back, offers another option.

Can I be unobtrusive? Ask people if you can watch where they go on the Web, and you are sure to give them the willies. But if you observe from a distance and then give them something in return, they are likely to thank you by coming back. "If you make it as easy as possible to get from Point A to the information they want, you're doing a good job," says Seamus McAteer, a senior analyst at Jupiter Communications. Wheelhouse's Ingari agrees: "I've never heard anyone complain about Amazon.com recommending books. On the other hand, everyone despises being called at home during dinner." The key here is to run fast without making it look like you're breathing hard.

Can I give my customers an out? Tell customers what you can do for them, but give them the opportunity to opt out if they're not interested--the "Thanks, I'm just looking" of the offline world. Letting that decision rest with the customer is crucial to the success of any personalization strategy.

RISKY BUSINESS

Even for companies that answer these questions affirmatively and decide that they and their customers are ready for personalization, a few caveats remain.

First comes the obvious issue of privacy. The prospect of all this free-floating customer data raises the hackles of privacy advocates and of customers who see something inherently voyeuristic about being watched while they surf.

But privacy incursions may ultimately be dwarfed by a much subtler danger--one that bedevils many adventures in technology: putting way too much stock in the technology itself and not enough in the underlying business sense that drives it. Personalization is no panacea. If your visitors are not going to open their wallets, there's no point in personalizing your business for them. You're running a company, not hosting a free party. So while it's important for visitors to appreciate their experience on the site, their level of enjoyment ranks second behind whether they click the "buy" button and become loyal customers.

Landsend.com offers female buyers an online mannequin they can build to their own measurements and use to try on clothes. That may make for a good time, but 80 percent of site visitors already buy from the Lands' End catalog. Does the addition of some admittedly cool technology mean people who traditionally don't buy from Landsend.com will do so now? Maybe not. But, wait.... If the virtual mannequin leads to more accurate sizing and, hence, fewer returns to process, maybe the ROI (to say nothing of customer satisfaction) goes up.

And for companies or industries that are not traditionally customer-focused, implementing personalization means a dramatic shift in business methods. "At the basic level, what's driving this is a change in the economic constraints in the way of doing business," says McKinsey's Singer. "The old constraint was the scarcity of space and bandwidth. Now shelf space is unconstrained; the real constraint is consumer attention."

Nike, for example, is a company that has historically competed and marketed largely on the basis of brand and image--much like all of its competitors in the footwear and sporting goods industry. When the company decided to offer visitors to its site the opportunity to build customized shoes, the move signaled a change in focus that may ultimately ripple through the industry.

Surely such a shift doesn't happen overnight, but it can be worth the exertion, since it has the potential to catapult Nike ahead of the pack.

But even after you've decided the investment is worth making, the sheer number of technology offerings can overwhelm anyone into a bad decision. "The challenge for marketers trying to do [personalization] is that there is an infinitely long list of things you can do to personalize," says Singer. "The real issue is where is the economic leverage."

No one product does everything (though many claim to), so the challenge lies in wading through an increasingly thick stack of vendor proposals to settle on something that delivers the right benefit for the right level of investment.

Companies that aren't clear about where the economic returns lie in relation to what they spend on personalization could end up in the poorhouse before even one customer has noticed their attentions.

How much is a lot to spend? According to Brad Allen, vice president of advanced technology for personalization vendor Be Free, that depends on how far you want to take your personalization strategy. To stick a toe in the water, says Allen, is likely to be a $500,000 investment. (Coincidentally, he notes, Net Perceptions and Wheelhouse ran a contest with a prize of $500,000 worth of personalization software and services; in June, OshKosh B'Gosh was announced the winner.) A company with big ambitions and intent on assembling its personalization capability internally can expect to invest "from single- to double-digit millions just to get into it," says Allen.

Be Free, whose offering is based on the ASP model, charges a flat monthly fee to "rent" personalization capabilities (fees start at $5,000 a month). Allen happily positions Be Free as the low-cost alternative to the kinds of homegrown infrastructure only multimillions can buy. "And you have to remember ongoing maintenance," he says, placing that burden at around "three full-time equivalents annually" for even a modestly scoped system.

It's also worth remembering that these technologies--and their adroit application--are still in their infancy, relatively speaking. Lynne Harvey, a senior analyst at the Patricia Seybold Group in Boston, thinks very few companies have actually achieved a "360-degree view" of the customer--defined as engaging in a relationship to learn as much as possible about all sides of the customer. "Some people think [personalization is] just a strategy of peppering any content with additional messaging" that recognizes the customer or adds a personalized greeting.

But it can and should be more than that.

Jupiter's McAteer agrees: "If Amazon.com is supposed to be state of the art, we've got a long way to go." Selling online is not just about personalized content, it's also about offering strong customer content and about making sure you can keep your customers coming back again and again. Fancy technology isn't going to do that on its own; it needs to be coupled with a customer strategy to build long-term loyalty. "Personalization is no substitute for love," notes McAteer--sage advice for companies trying to balance customer needs with economic realities and the promise of technology. And while it might make for a pretty good one-night stand, a company lacking the right vision and a strong customer commitment could end up disappointing someone in the morning.

Got a beef about personalization? Send it, along with your name, SSN and all of your bank account information, to Senior Writer Meg Mitchell at mmitchell@darwinmag.com.

CASE STUDY BMW OF NORTH AMERICA www.bmwusa.com Industry: Automotive Goal: To provide a personalized community for BMW owners that secures their loyalty; to convert nonowners into first-time buyers. The customer gives: No information unless he wants to be contacted by a dealer. Owner's Circle members cough up name, vehicle identification number and date of purchase.

The customer gets: To build cars to their own specifications based on information about model types and driving preferences, then can save preferences to return to or ask to be contacted by a dealer. BMW owners can join the exclusive Owner's Circle, which lets them track scheduled service and receive mileage reminders and other information about the care and feeding of a BMW.

Stickiness factor: BMW owners are notoriously loyal to their cars; the hope is that they will feel the same way about the site and return often. And new buyers who might be intimidated to shop for their first Bimmer through a dealer can do the legwork on the Web.

What about privacy? Site visitors can opt out of anything; BMW promises never to sell its coveted customer list to anyone.

Challenge: To use personalization in a manner that doesn't offend customers who are traditionally opposed to a hard sell; to make the site personal but not intrusive.

CASE STUDY KAPLAN/ESCORE.COM www.escore.com Industry: Education and training Goal: To garner lifelong customers for Kaplan's games, educational products and learning activities.

The customer gives: Name, e-mail and basic information about their child, including educational or developmental information, in order to register.

The customer gets: Profiles of children and offers for relevant products and services. Parents can take tests to assess their parenting abilities and can direct specific questions to experts.

Stickiness factor: If Kaplan can convince parents the site will help them in different stages of child rearing and sell them the right products at the right time, they'll want to come back. After all, it takes a long time for a kid to grow up.

What about privacy? A very touchy matter when it comes to children. Kaplan shares aggregate information with outside sources but does not attach names to it.

Challenge: Using the Web to reach across age segments, broadening the image of Kaplan as more than a test preparation company.

CASE STUDY GREATCOFFEE.COM www.greatcoffee.com Industry: Gourmet coffee retailer Goal: To turn first-time browsers into buyers.

The customer gives: Nothing. The company uses Angara's E-Commerce Targeting Service, which delivers variable content to visitors based on their geographic location.

The customer gets: Homepages personalized to his area (a visitor from Taiwan might get information on international shipping; one from the Bay Area might get a chance to win San Francisco Giants tickets).

Stickiness factor: The industry average for reordering within six months is 5 percent; GreatCoffee's is 63 percent. Plus, coffee is addictive.

What about privacy? Customers don't need to give up info until they're ready to buy; once they do, the company maintains a strict privacy policy and insists that its partners do the same.

Challenge: Focusing on customer service after the first purchase to keep reorder rates close to conversion rates.

SOME TERMS EXPLAINED A glossary to help you navigate the roads of personalization Collaborative filtering: Software that aggregates customer information and makes recommendations based on the preferences of like-minded customers Rules-based engines: Use Boolean logic to map customer profiles to content so that customers can get specials based on factors like where they live or what they bought last time Cookies: Computer files that assign anonymous numbers to Web users, making it easy to track their movements around a site. Site administrators aggregate and analyze cookie information, which can be useful in evaluating site effectiveness as well as in identifying and greeting repeat visitors. It is possible for individual users to configure their browsers to disable cookies, but few take the trouble to learn how to do this.

Clustering: Observing customers' online behavior and placing them in similar behavioral groups, then inferring what a group member will do or buy next based on the retail behavior of the other group members Sticky sites: Sites that manage to keep customers or get them to come back once they've left Data brokers: Third-party companies that collect and disseminate customer data, culled from a variety of sources like software registration and publicly available information Real-time personalization: Reacting to a user's behavior as it happens rather than going back and analyzing it later Opt-in personalization: Letting the customer say yes or no to letting a company use personal information, with the opportunity to opt out at any time Data warehouse: A store of data that has been collected and organized for the purpose of being analyzed Data mining software: Software that discovers previously unknown relationships among data PRIVACY, SCHMIVACY A lot of people have a lot to say about privacy on the Web.

But is anybody really listening?

An April 2000 report released by Cyber Dialogue, an Internet customer relationship management company in New York City, found that 21 percent of Internet users don't know if their browsers are set to accept cookies or not.

Even so, 71 percent offer up personal information in exchange for personalized content. The survey also found that more than 80 percent of users don't mind providing personal information such as name, level of education, age or hobbies; only 59 percent are willing to share their household income; and a mere 13 percent feel safe divulging their credit card numbers.

But privacy advocates still quake at the notion that Web consumers leave a lot of information behind that they might not even know about. For Jason Catlett, president of Junkbusters Corp., a Green Brook, N.J.-based company that keeps a close eye on privacy infringements, it's a simple case of mathematics. "You can only fill out so many warranty cards in a year, but you can leave thousands of electronic footprints behind in a day," he says. Most people, he adds, do not even understand what cookies are or what they do. And while sites that rely on cookies to collect user information claim that they preserve visitors' anonymity by assigning them a random number, once a browser becomes a buyer and pulls out a credit card, the cookie number gets a name attached to it.

Then--zing!--the marketers move in.

One answer to the privacy conundrum, already adopted by many companies, is to write and post a privacy policy, letting customers know exactly where their information is headed. "But most privacy policies read more like privacy disclaimers. They want to be deliberately vague so that the companies can sell info without violating the policies," says Catlett.

Another solution--most fervently advocated by companies with a stake in unfettered data collection--is self-regulation. In April, 26 companies came together to form the Personalization Consortium, an advocacy group with the goal of promoting responsible use of technology. Members include American Airlines, personalization software vendor BroadVision, Peppers and Rogers Group (a marketing consultancy) and PricewaterhouseCoopers, among others.

Sun Microsystems CEO Scott McNealy has another idea. McNealy made headlines last year with this now-famous remark: "You already have zero privacy--get over it." Chances are that many Web customers have probably already heeded his advice, but it's early yet. Buried beneath the e-commerce battlefield, plenty of privacy land mines lie in wait. And companies better step carefully. -M.

Mitchell

CASE STUDY PLANETRX.COM www.planetrx.com Industry: Health care Goal: To capture the burgeoning online health-care market by replacing the old-fashioned pharmacy on the corner with an online service that dispenses advice and wisdom as well as prescriptions.

The customer gives: As much as he wants: health and insurance information for prescriptions, less for general over-the-counter shopping.

The customer gets: E-mail reminders for prescription refills, access to online pharmacists and a secure, password-protected environment in which to receive answers to personal questions. Visitors can also gain access to specific disease sites with communities for caregivers and patients.

Stickiness factor: Loyalty is likely; customers won't want to impart sensitive health-care data to too many parties, so they'll stick with the company that does it right.

What about privacy? Of the utmost importance: "Privacy is our lifeblood," says Matthew Naythons, publisher and vice president of editorial.

Challenge: Safeguarding some of the most sensitive data in the world and overcoming customers' natural resistance to talk health with someone they can't see.

Join the newsletter!

Error: Please check your email address.

More about AdroitADVENTAmazon.comAmerican AirlinesAMR ResearchBMW Group AustraliaBroadvisionCyber DialogueForrester ResearchImpartJupiterJupiter CommunicationsKaplanNet PerceptionsNikeOshkosh B'GoshPersonalization ConsortiumPlanetRxPricewaterhouseCoopersPricewaterhouseCoopersRogers GroupSeyboldSun MicrosystemsWheelhouse

Show Comments

Market Place