Merrill Lynch Phones Ahead

The Wall Street giant is making a major bet on Internet-based telephony as a way to improve service and enhance flexibility. Here’s a case study on the promise and pitfalls of technology-driven innovation.


Talk about a wake-up call. John McKinley picked up his office phone and discovered that the line was dead. For most executives, this would be a nuisance. For the CTO of Merrill Lynch & Co., it was a potential disaster. A phone outage can mean millions of dollars’ worth of lost business for the largest brokerage firm in the United States.


Fortunately, McKinley had another telephone on his credenza, a new device that uses the Internet and Merrill’s high-speed data network to handle calls. The CTO had tinkered with the new phone, which was designed by Cisco Systems to take advantage of an emerging technology called “voice over Internet protocol” (VOIP), but McKinley had never relied on it to do business. When he lifted the handset, he heard a dial tone and got to work. For the next two hours, he used it to troubleshoot the problem, which turned out to be minor, until AT&T restored service.

When it was over, McKinley regarded the device with newfound respect. Merrill has big plans for VOIP, and this trial by fire showed McKinley that VOIP could pass the crucial first test: It worked like a normal phone. “The phone is the most important tool in our business,” he says. “It’s the only service that’s 99.9% reliable. But no one is going to praise a system for having a dial tone. It has to be so reliable that people aren’t even aware of what device they’re using.”

Big Possibilities, Immediate Payoff

A couple of years ago, Merrill looked like dead meat as it scrambled to respond to the threat posed by competitors such as E*Trade and Charles Schwab. But the firm emerged from that scrape as a born-again innovator with a potent Internet presence and a renewed commitment to applying technology as a competitive weapon. The VOIP phones are the stealth bombers of the company’s arsenal.


About 3,000 of the company’s 68,000 employees have the new phones now. McKinley plans to install VOIP phones in every new office starting in 2002, a target to which Merrill remains committed despite tough times and retrenchment on Wall Street. His case for the new technology is rooted in the belief that Merrill can’t prosper by cutting trading costs to the bone. The firm lives and dies by customer service, easy accessibility, and global reach. “We are a high-touch business, and we sell bits, not atoms,” says McKinley. “So adroit use of technology can be a way to gain tremendous competitive advantage.

“Imagine a phone that recognizes a customer before the broker even picks up the handset,” McKinley continues, “and flashes a snapshot on a computer screen of the state of that customer’s portfolio, along with his 10 most recent trades and how those stocks are performing. Such applications embody VOIP’s potential — a chance to improve our productivity and responsiveness to customers in deep, rich new ways.”

But an emerging technology can’t pull its weight at a place like Merrill if it doesn’t also deliver a quick payoff. To improve the odds of successful early adoption, McKinley mapped out areas where VOIP could help Merrill save money fast. Indeed, before he’d even plugged in the first VOIP handsets, he saw a way to cut Merrill’s phone bill. “I’m paying for the voice circuits to carry phone calls around the block or around the world,” McKinley says. “And I’ve paid again to put in this great data network, which has grown huge in order to handle this gusher of information. Can’t I have some convergence and start moving my voice traffic over the data network?”


VOIP technology can do that because it exploits a key difference between the network that has evolved to handle voice calls and the network that has been built for data. When you speak into the phone, your voice is turned into an electronic signal that is then diced up into discrete segments called “packets.” Those packets are labeled with electronic tags and fed into the phone network, where they blend with millions of other voice packets. When they reach the other end, the packets are reassembled and sent to the phone of the person you are calling.

Internet-protocol networks also use packets to move around discrete bits of information. But unlike the traditional phone system, which opens up a dedicated line or circuit between your phone and the phone that you are calling each time you place a call, IP traffic takes the fastest route available. Different packets of data from the same file or email message may travel north through Minnesota and North Dakota, and south through Texas and Arizona, to get from New York to San Francisco, where they are reunited.

That’s fine for data, where a few milliseconds of lag time on a transcontinental journey is no big deal. But voice calls are more exacting. The slightest delay in reuniting the packets can … make … a phone … call … sound like … this. It’s annoying if you’re trying to talk to your college roommate. It’s fatal if you’re doing millions of dollars’ worth of business each day over the phone.


Merrill’s Most Demanding User

Jerry Curtis had been living with that annoyance for a year, and he was starting to run out of patience. Curtis is the prickly conscience of Merrill’s technology strategy. Tall and rangy, with a beard and shaggy hair, he looks more like the captain of a charter sailboat than the guy who is responsible for making sure that the computer systems used by Merrill’s brokers don’t crash. He has the right temperament for the job: brutally honest, self-deprecating and smug by turns, and he’s not shy about using his power to insist on changes. “I’m not the greatest visionary at Merrill Lynch,” he says. “I’m considered ultra tactical.”

In his role as a technology advocate for Merrill’s stockbrokers, Curtis travels regularly to different offices and watches how the brokers actually work when they’re on the phone with customers. “It gives me a different perspective,” he says. “I’m really a businessperson who just happens to be marooned in technology. I do email, and I’m technology literate. But I know that the people who I serve are not.” That’s why Curtis has the least powerful computer at Merrill Lynch on his desk, along with a dinky monitor. His throwback equipment lets him spot the shortcomings of new software applications or network upgrades that have been built and tested on state-of-the-art equipment.

Curtis’s “ultra-tactical” perspective was in short supply when Merrill started experimenting with VOIP in early 2000. From the moment he picked up the receiver, Curtis found fault. Why, he wondered, would answering or placing a call from the new phones require a two-step process? First you had to lift the handset, and then you had to push a button to be connected or to get a dial tone.


When you stopped talking, the phone would be dead silent. Curtis kept thinking that he had lost the caller, and he constantly found himself saying, “Are you there?” Sometimes the VOIP phones would drop the connection midsentence. Curtis’s fiancée in Florida hated the way the prototype phones would clip off the tail end of the last word in a sentence in order to optimize bandwidth instead of voice quality. She would tell Curtis to call back on a real phone and hang up on him. “It was a bad experience,” he says. “They came up with a phone that didn’t behave like a phone.”

Curtis’s list of grievances was long and detailed, and his complaints were obvious — which made him a great candidate for the test team. The guys down the hall wanted to reclaim his VOIP phone while they worked on the problems, but his boss vetoed that move. Curtis was stuck with the new phone. He quickly became Merrill’s most demanding VOIP user.

Interesting things happen when you give an emerging technology to a powerful nonconformist and watch what he does with it. Curtis started carrying the VOIP phone around with him to different parts of the sprawling Merrill Lynch campus in Hopewell, New Jersey and plugging it in to spare data wall jacks. When people called his phone number, it rang through to wherever he was.


He carried it to Florida once, sneaking it out of Merrill’s offices in his backpack at the end of the day, just to see how it would work from home. It took a while for the phone to make a connection and to register itself as an active number in the Merrill directory, but eventually he got a dial tone. The voice quality was rough, because Curtis was relying on a slower IP-network connection from home. Still, he was impressed.

Cisco’s engineers worked with Merrill to address Curtis’s other concerns. They injected the faintest hum into the phone when a call was active, a “comfort noise,” so that people would know that they were still connected. The voice quality was groomed to avoid clipping words. When Curtis calls his fiancée these days, the only way she can tell that he’s using the VOIP phone is by looking at the caller ID.

Eventually, Curtis got his regular phone back. Far from being a VOIP skeptic, he had turned into a champion of the new technology. “This is a voice business,” he says. “But we’ve been using a device that hasn’t advanced very far since the day it was created. Now that we have a new set of features that no one ever imagined, we can start to do a lot of things differently.”


The Lure of “Silicon Economics”

The promise of doing a lot of things differently is what most intrigues John McKinley. About a year ago, he pushed Nortel Networks to supply an upgrade to Merrill’s conventional telephone-switching equipment so that voice and data could share the same network. That efficiency lopped 25% off of Merrill’s phone bill. The next generation of VOIP switches, from Cisco and other suppliers, gave McKinley even more ammunition for change.

Because the new VOIP systems are built on Microsoft’s Windows NT, they are simpler to manage than the proprietary telephone equipment that they have begun to replace. Their open architecture provides a silver bullet, for instance, to a logistical nightmare in business: the cost of changing phone numbers when an employee moves. Merrill pays the phone company from $100 to $150 to move, add, or change a number, which happens about 20,000 times a year.

“With VOIP, you unplug your phone and put it in a box with your family pictures,” says McKinley. “When you get to your new location, you simply plug the phone into the wall, and it registers itself.”


Before the Internet and telecommunications crash, such game-changing possibilities were admiringly known as “silicon economics”: the promise of a huge leap in price-performance ratio, thanks to combinations of hardware and software that are based on common standards. It is the same principle that allowed PCs to push mainframe computers to the margins and put cell phones into the hands of millions. It is also the concept that launched a thousand dotcom business plans — which is one reason you don’t hear the term much these days. But silicon economics is alive and well in the form of VOIP, and it’s following its customary pattern of disruption.

The heart of McKinley’s business case for VOIP — the big vision of what it will do for Merrill — comes down to a bet that silicon economics will prevail.

The possibilities are truly dazzling. McKinley sees VOIP as a way to push key messages out to Merrill’s troops: “We can flag our financial advisers, sales force, and senior management that a star tech analyst at Merrill just gave a talk about the latest developments in the instant-messaging war between AOL and Microsoft, for instance. Hit this button to tap into a replay of the call.”


Eventually, when VOIP is picked up by banks, mutual-fund companies, institutional investors, and other customers, McKinley believes that the new technology will be a means for even faster, broader distribution of Merrill’s intellectual property.

“We’re a huge financial publisher, if you think about the eight hours of TV content and the couple of linear feet of research reports that we create every day,” he explains. “So we have to start thinking like a new-media company, and one way to do that is to use VOIP as a platform. We can use it to call people’s attention to our perspective on the financial markets and let them tap into Merrill’s bit stream of information.”

Merrill is hard at work on another application of VOIP that offers a tantalizing glimpse of its future. It will give McKinley the means to turn on a dime and reconfigure Merrill’s operations even for short-term sales campaigns, something that would have been prohibitively expensive in the past. “When we send out a mailing to our customers about the new tax laws, for instance, we expect a spike in the number of inquiries from them,” he says. “Using VOIP, we’re able to spread the load of those calls to tax-specialist brokers around the country without breaking a sweat. That’s very powerful. Ultimately, in today’s world, the easiest companies to do business with will be the ones that win.”


Paul C. Judge ( is a Fast Company senior editor. Contact John McKinley by email (

Sidebar: Speed Dial

If you haven’t made a phone call yet using voice over Internet protocol (VOIP), you will soon. Here’s why.

Convergence at Last: Because VOIP uses the same standards as data communications, voice calls, email, and even video broadcasts can go through the same network. People can get voice mail on their PCs and check email with their phones.


Wherever You Go, There You Are: Using VOIP, people can log in from anywhere in the world to receive calls and get access to their own set of phone services. All it takes is an Internet connection. That degree of flexibility provides benefits across the board, from telecommuters to fast-growing operations overseas.

The Next Best Thing to a Digital Toaster: VOIP phones may be the first true information appliances: simple to use, loaded with power, and capable of opening a window to the Web with the push of a button. Merrill Lynch, for example, has quickly linked its VOIP phones to applications developed for the Web, such as a phone directory that searches for employees with certain skills.

Let a Thousand Features Bloom: Early adopters of VOIP are building their own applications, but the open architecture of VOIP is enticing dozens of companies to come up with programs to extend the phone’s capabilities.