|
Data mining goes online
|
 |
September 24, 1999: 12:04 p.m. ET
Online brokers want to get to know you, track you and give you what you want
By Staff Writer Alex Frew McMillan
|
NEW YORK (CNNfn) - Are you fed up with your broker calling during dinner to see if you want in on a hot stock? Or a little paranoid about privacy? If so, you might not like the idea of data mining. Like it or not, if you trade online, you're almost certainly going to see the effects of data mining over the next few years, as online brokerages use that technique to target their customers.
The main online brokerages are taking a page out of the book of the banking industry, which already uses data mining extensively. A financial institution tracks information on its customers -- gleaned from what they tell it, what they do or by buying data -- and then puts it in a computerized data warehouse. The simplest version lets a company know where their customers live so it can target an area to send mailings to or advertise in for new customers.
But increasingly, brokerages are using sophisticated software to work out who buys what and why. That more advanced data mining is sometimes called "decision mining." By building as specific a profile of each customer as it can, a brokerage, or any kind of company for that matter, can map that against the products different customers tend to order. Ultimately, with the right modeling and software, they get a good idea of how different people act, and why.
Cross-selling opportunity
With many online brokerages expanding into areas like banking or broadening their range of products, there's a good opportunity for some to "cross sell." And the Internet is an excellent venue for gathering that type of data, which explains why online brokerages are keen to make the most of what you're clicking on.
"It's an electronic mechanism, and we know that you like to buy Microsoft at 3 in the morning," said Rob Bethge, chief marketing officer at the brokerage Datek Online. "There's the ability to find out what stocks you got news on, and then what stocks you buy or sell. Or that someone got 500 quotes before they placed a trade, or that every time they got a quote they made a trade."
Maybe you like mutual funds. Or maybe you keep logging in and keep using an asset-allocation tool, plus you've got kids who are approaching college age. Perhaps you've jumped into an IPO or two and there's another coming up. Or you just love those Internet stocks and an influential tech report is coming out.
Like offline brokers who try to develop a relationship with their customers, online brokerages hope they can learn what you want and give it to you, perhaps before you even ask for it. Only online brokerages think they can do it better, with more reliable information and less obtrusive ways of reaching you. It'll be a fax, page or, most likely, an e-mail, instead of a phone call interrupting you over your meat loaf.
"It's a lot less intrusive than calling you at home during dinner and saying 'Gee, is this something you'd be interested in?' " said Blake Darcy, chief executive of DLJ Direct. His company is developing its data-mining software to offer specific, personalized e-mail service. That may take the form of tracking what particular stocks you trade and "pushing" information to you on those issues and others like them.
Darcy is using book seller Amazon.com's customized service as his guide. "Everyone thinks that the nice service about Amazon.com was, you didn't have to do anything about it and it happened."
Perilous field
But he also knows data mining is a perilous field to traverse. "There are probably going to be people who say 'I don't want unsolicited e-mails this way,' " he admitted. DLJ Direct (DIR) probably will provide an option for customers to opt out, though the company isn't that far along.
The Securities and Exchange Commission also is worried that by forwarding information to customers, online brokerages are getting in the business of recommending stocks. Online brokers are leery of taking on "suitability," an obligation of brokerages that recommend investments to track whether those recommendations are suitable to investors. It opens them to claims from investors that they were unsuited for an investment that turns bad.
SEC Commissioner Laura Unger says that has yet to be hashed out. "If we interpret suitability too broadly, then we will limit the information that investors receive, and that's not what we want to do. If we interpret suitability too narrowly, we gut the intent of the suitability provisions," she said. DLJ Direct's Darcy said his company will work with the SEC to iron out what the brokerage needs to do.
Fear of misuse
Customers' biggest concern is that their brokerage will misuse the information their brokerage gathers. Though they liked Amazon.com's service, customers were put off when it started disclosing the most popular books at particular companies, for instance. Even aggregated information like that "gets people nervous," Darcy said.
The biggest fear among customers is that their brokerage will sell the data they've mined to an outside company, a mistake several banks have made. "If you do that, you're dead," Darcy said. "You're giving very sensitive financial information to an institution for a reason … and it's not so they can take that information and profit from selling it to someone else. It's a clear misuse," he said.
Though the tendency may be for customers to shy away from the idea of data mining, the industry expects the personalized service it leads to to catch on. Carlos Bhola heads investment bank CS First Boston's Internet and electronic commerce group, which works with the majority of the big online brokerages. He thinks big changes in the industry are "imminent."
"These systems do not only offer threats. There are some significant benefits -- value creation, much enhanced customer satisfaction. I think they should be viewed as such. Not just the Big Brother, 1984 syndrome," Bhola said. Bottom line: customers may well like getting an inside line on an IPO offer in their favorite industry, or in feeling like their brokerage really knows what they want.
Charles Schwab (SCH) and Waterhouse Securities (TWE) have been the forerunners in understanding their customers through technology. E*Trade (EGRP), Ameritrade (AMTD), Fidelity and smaller brokerages such as National Discount Brokers are, like DLJ Direct, developing data mining capabilities.
Seeking 'wallet share'
The driving force, Bhola said, is that online brokerages don't want to compete just on price any more. They want to retain customers, many of whom have been fickle and opened numerous accounts. And they want greater "wallet share" of financial services, attracting customers to insurance, banking and loans.
"These guys in the online space are now figuring out what are you reading before you bought a stock," Bhola said. "On the Web we can see you read the article on Goldman's earnings yesterday, clicked through to the Goldman site to understand it a little better, then you came back and bought five shares of Goldman."
Only the low-cost, transaction-oriented online brokers such as Datek won't be drawn to data mining, he thinks. And it's not because they can't do it. "The difference isn't really capabilities. It's interest." Datek's Bethge confirmed the company could easily mine for customer data, but Datek doesn't really do much marketing, focusing on fast, cheap execution instead.
Of course, brokerages have long tracked which areas of their sites get the most attention or aren't used, and whether there are dead ends in their sites. But they're only now starting to focus their attention on tracking more specific, individualized information.
Relationship management
Over the next five years, the securities industry will boost its spending by 14 percent a year on the customer-relationship-management software that lets them mine for data, according to Meridien Research Inc. That's faster than any other area in financial services. By 2003, brokerages will be spending between $250 million and $300 million a year on such software systems, with the Internet a key focus.
Expect to see a big push to target YOU start next year. "It's a year 2000 event," Darcy said. "This year it's wireless trading and after-hours trading. Next year it'll be the computerized advice, customized service."
Though customers may balk at the idea at first, the brokerages think most consumers will come to demand the "personalized" attention and Amazon.com-style customer service that data mining allows.
Michael Anderson, vice president of investor and public relations at Ameritrade, said he thinks data mining will get a lot of attention over the next two to three years, both from customers and the companies that serve them.
His company is working on how to tailor its products for its customers and expects to roll out customized service over the next few months. The Amazon model is something the company has "kicked around," though Ameritrade is not sure how to apply it.
"We all will be looking at how to build a relationship. The more we're able to use technology, it allows me to be more efficient and allows me to keep my cost structure down and provide better service," he said.
And customers will demand it, he thinks. "It's the world of atomized marketing," he concluded. "You'll have to provide that."
|
|
|
|
|
 |
|
