It was every consultant's nightmare. Abdou Touray was on assignment late one Friday afternoon. He was on his own, and he was stumped. A major mutual-fund organization had placed an urgent call to Touray's company, Context Integration, because its computer system kept crashing. Its executives were desperate: Investors were counting on them to manage billions of dollars in assets, and if the system kept crashing, its fund managers couldn't manage anything.
Context had been courting this company for nearly a year, so this was its big chance to impress a potential client. But this was no easy task. About 30 in-house techies hadn't come up with a solution.
Now it was Touray's turn. Touray, 28, may have been working on his own that day, but he wasn't really alone. Supporting him was a Web-based knowledge-management system. Around 5 am on Friday, Touray fired off a "911" message from his laptop, and the system automatically contacted Context employees who are experts on application servers. Over the next hour, colleagues working in New York City, San Francisco, and Houston responded to Touray via email, creating a discussion thread to wrestle with the situation. Finally, something clicked. Touray realized that a URL being relayed from a browser to the system's server was too short. That simple error was causing the system to shut down. Problem solved.
"I let the others know right away," Touray says. "And I'm sure that the news made their weekend." The news certainly made an impression on the client. First, the mutual-fund company's executives exchanged high-fives. And soon afterward, they hired Context for a major new project.
So it goes in the new world of business. Winning companies don't just outhustle or outmuscle the competition. They out-think the competition. Business today is about brains, not brawn. It's about how many ideas you generate, not how many factories you own. And ideas come in many shapes and sizes. Every so often, a company will invent a breakthrough business model that reinvents the rules of competition in an industry. But there's a day-to-day side to competing on ideas: Can your marketing people in Seattle quickly make use of a presentation that wowed a client in Savannah? Can a programmer with a problem in Los Angeles quickly tap the expertise of colleagues in Austin? Can a mid-level executive in Chicago toss out a proposal for a new line of business — and trigger a companywide conversation that turns the idea into something real?
Context Integration is a fast-growing Web-solutions company based in Burlington, Massachusetts. It has compelling answers to all those questions, and the answers start with a $10 million knowledge- management system — a computer network that stores best practices, tracks new ideas, fields questions, and operates as a kind of group mind. It is the killer app in an organization that understands that the most powerful weapon in business is a great idea. "We have positioned ourselves on the frothing part of the knowledge wave," explains Bruce Strong, 43, vice president of strategic services, who is one of the company's four founders. "If we're not sharing ideas with one another, with our clients, and with our partners, we're in deep trouble."
Context is in anything but deep trouble. The company employs 200 consultants, all trained to tackle the most demanding challenges in the world of e-commerce technology. They build high-volume Web sites that enable companies to sell products, to trade commodities, and to manage financial risk. And they build those sites fast. A Context project relies on short, intense development cycles (each lasting three to six months) and comes with a fixed price (usually less than $1 million). Every year since 1993, Context has doubled its annual revenues. This year, its staff will nearly double as well, to about 350 people. That's the power of competing on ideas.
To be sure, plenty of companies have built electronic repositories of best practices and customer information. But most of these databases involve the electronic equivalent of musty library shelves. Context's system, on the other hand, functions like a central nervous system — a system that connects staffers working in eight offices and in a couple dozen client sites nationwide. As consultants execute their projects, the network steers them to resources and experts that might be of help. Every day, it distributes articles and tips to employees who might find them relevant. When a problem arises, the network tracks down the people who are best qualified to solve it. "Lots of companies try to do knowledge management and then fail — because they don't understand that knowledge management has to be an enterprisewide priority," says Pete Solvik, CIO of Cisco Systems and a Context board member. "At Context, knowledge is a core strategy. It's embedded in the culture."
No One Is as Smart as Everyone
There are times when Context Integration sounds like a one-person company. The same name keeps popping up in conversation. "We couldn't possibly meet our dead- lines without Ian," says one Context staffer. "If I don't know how to solve a problem, I just ask Ian," says another. "Once we show our clients what Ian can do, they're sold," says yet another. After a while, this Ian starts to sound like everybody's favorite colleague — indeed, like an absolute genius.
In fact, Ian is IAN, the company's Intellectual Assets Network. A Web-based knowledge system that Context built from scratch in 1997, IAN was the brainchild of Bruce Strong and Stan Ward, 33, the company's chief methodologist. "We wanted to be better than our individual selves," says Strong. If individual consultants could tap into what other people in the organization knew, they could work smarter and more productively. They could help one another on technically challenging projects, and, instead of solving the same problems over and over again, they could free themselves to attack new ones. So Strong and Ward developed IAN, which functions as a kind of group brain.
Context, founded in 1992, with offices stretching from the East Coast to the West Coast, had long seen collaboration as a challenge. Lotus Notes had been helpful at times, but it never became the indispensable "go-to place" that Ward envisioned. So he and Strong decided to take matters into their own hands. But first they embarked on a "world tour," visiting every Context office to find out exactly what the company's consultants wanted in a knowledge-sharing system. Everywhere they went, they got the same feedback. They heard, "I have to be able to use it anytime I want to" — so the two made sure that IAN was available on the Web, as well as on every laptop hard drive. They heard, "It has to be fast and convenient" — so they created a search engine tailored to Context's homegrown vocabulary. They heard, "The content has to be accurate, up-to-date, and clean" — so they hired someone to serve full-time as a moderator for IAN.
That someone is Mary Durham. A cheerful and energetic midwesterner, and one of many Sybase alumni who work at Context, she is an exacting knowledge manager who works in the company's Boston office. Durham, 47, says the real power of IAN lies in its simplicity. She compares the system to her favorite kitchen appliance — a mixer — because of its all-around utility and its absolute ease of use. When a project team begins a new assignment, for example, IAN walks members through the development process, offering sample templates, sample code to borrow or modify, and detailed descriptions of each step. "This seamless interconnection gives team members the feeling that someone has thought of everything — on their behalf," says Durham. "They don't have to hunt for what they need."
At Context, it has become virtually impossible not to use IAN. "It's part of the ether," says Strong, who works in the San Francisco office. The IAN home page resembles a solar system, with subject areas — "Project Artifacts," "Knowledgebase," "Discussions" — serving as planets. In "Partners," consultants draw on a cross-referenced listing of each business partner's in-house experts. In "Colleagues," project leaders learn about new team members: their professional specialties, their technology interests — even what they look like (most entries include a photo).
Yet IAN's real value lies not just in the ideas and information that the system stores, but in how that material affects work in the field. During another assignment, this one for the Bermuda Commodities Exchange, Abdou Touray was trying to fix an Internet trading system when the local phone lines went down. Touray was cut off from his colleagues and from the rest of the world. But he still had IAN — in the form of an updated copy on his hard drive. In just five minutes, he found what he needed to fix the client's system: a rather complex algorithm that might have taken him a week to write. Fortunately, another consultant at Context had encountered the same problem a couple of years earlier. That consultant was no longer with the company, but his expertise was. "It's an eerie feeling when you see knowledge captured like that, especially when you're isolated on a job," says Touray. "It's almost as if you're being haunted, as if someone is reaching out from the past to help you solve your problem."
Still, group intelligence is about more than solving specific problems. The real power of sharing knowledge, says Stan Ward, comes when an entire group is able to build on an idea — to follow a thread of innovation wherever it leads. Ideas are unpredictable, fleeting, and hard to recapture once they're lost. That's why Ward, for all his love of technology, still carries a well-worn, brown leather notebook, in which he jots down whatever comes to mind. "If I leave my notebook somewhere, I'll ask myself, 'Where did I leave my brain?' " says Ward, who has saved nine years' worth of notebooks.
Virtually speaking, IAN lets Ward, and everyone else at Context, open their notebooks — so that others can react to the information recorded there. IAN serves as a marketplace of ideas. It's a digital brainstorming center in which all ideas, even half-baked ones, are welcome. After all, it's hard to predict which suggestion will trigger a chain of thought that leads to the next big thing. "It's that interplay with other people that helps an idea take a quantum leap," says Ward. "We want to foster that kind of communication and to capture it as it happens." That's why people are encouraged to ask IAN a question — any question, whether or not it's urgent. IAN eavesdrops on the conversations that such questions generate, records participants' input in a discussion thread, and then disseminates that knowledge to people who want it.
In other words, IAN doesn't wait for you to come to it — it comes to you. The system is "in your face," Durham says. Each morning, for example, customized electronic newsletters go out to consultants who subscribe to them. The system distributes only those kinds of ideas and information that users ask for. This approach avoids information glut, which Durham regards as one of "the enemies of knowledge management."
Newsletter topics range from particular technologies, to Web-development platforms such as BroadVision and SilverStream, to colleagues whose work a consultant wants to follow. For example, if a consultant is interested in corba (common object request broker architecture), she will receive a message linking her to the latest corba material that IAN has generated. This newsletter might refer to a trade-journal article, to a question posted by a colleague, or to a posting by Tony Lanzilloti, 39, a technology manager who is an expert in corba. To read a question, subscribers must follow a link to the discussion area. Why the extra step? To encourage users to reply in IAN, rather than by phone, so that their ideas can be captured and shared. "They understand right out of the gate that IAN is not all about 'take, take, take,' " Ward says. "There's an individual responsibility, a social responsibility, to give back."
Do You Know What You Know?
Competing on ideas is an extremely demanding way to work. Companies, and the people in them, are expected to generate cutting-edge innovations at a steady clip, to execute those innovations flawlessly, and then to produce quick results. Indeed, one of the defining questions in business is "How do we do great work fast?" One of the best answers is "Wherever possible, avoid reinventing the wheel."
IAN is an idea-recycling center. Teams at Context begin every new project by searching IAN for past work that's similar to what they're doing, looking for ideas and practices that they can borrow. The goal is to use anything — models, architecture, source code — that helps them avoid duplicating another team's work.
Last year, when Context approached BMG Distribution, the $4 billion subsidiary of Bertelsmann AG, it didn't waste any time making an impression. BMG wanted to build a Web site to handle sales and marketing activities with distributors and retailers across the country. It was no small undertaking: BMG moves more than 600,000 units daily to music, CD-ROM, and video retailers. Context was a latecomer to the developer-selection process, but it promised to deliver a proposal within 48 hours. Its consultants conducted an intensive workshop with BMG's information-technology and marketing people, to learn more about what BMG wanted the site to do. Then, after combing through IAN, they produced a proposal that would otherwise have taken weeks to prepare.
"You could tell that they had put a lot of thought into the proposal," says Anthony Marino, project manager for applications development at BMG. "They didn't just show a slide or two. They had documents: 'This is how we're going to give you what you want. These are the people we're going to assign to it. This is why we can do the job.' " Last January, Context was in San Diego when BMG Distribution introduced its new Web portal, BMG Central, at its annual convention — right on schedule.
Open Minds Beat Closed Systems
Context is in the knowledge business — but it doesn't pretend to have a monopoly on good ideas. There are plenty of consulting firms and software-development companies that go to clients with an attitude that says, "We're smarter than you." Why else, they figure, would a client hire them in the first place? Context exudes a different attitude. "When we do a project, we don't do it to you, we do it with you," says Shahid Khan, 27, a client manager who runs Context's media practice. Adds Dror Liwer, 31, the firm's New York - practice director: "We've had clients ask us not to use the C-word — 'consulting' — because it has such a negative vibe. There can be tremendous distrust, because consultants usually hide so much from their clients."
Context organizes its projects as open systems — as construction sites unencumbered by walls or fences. Clients watch and critique the iterative Web-development process, observing each step via the Project Homepage — a password-protected Web site on which consultants post a project's latest status report, along with lists of bugs and risk factors. "There's nothing wrong with letting a client see our mistakes," says Liwer. "This is all about trust, and trust is built through sharing strengths and weaknesses."
This open process works for Hearst Distribution Group, which recently hired Context to build its book-ordering solicitation system. The result is one of the first sites to let publishers view orders in real time — which means that they can quickly increase a print run or alter a book cover, depending on sales figures. Nolan Bennett, senior vice president of the book division at Hearst, checks the Project Homepage daily. "You can see everything right away — the testing, the conversations back and forth," he says. "The Context team solves problems in a matter of hours."
Armed with laptops, Contextians capture a client's feedback in meetings and then transcribe it directly onto the Project Homepage. One client dubbed the process "Borging," because it reminded him of the alien race in Star Trek that assimilates entire cultures. Context likewise tries to absorb everything — in part to improve its work on individual projects, in part to increase the company's collective intelligence. (Unless a client specifies otherwise, all Context teams have access to every Project Homepage, so that they can review their colleagues' work and borrow material for their own projects.) Of course, "the Borg is evil," says Ward, who nonetheless ordered Borg figurines for his project team. "We're just trying to leverage 100% of everything that's available in the company."
On every project that Context undertakes, the initial wave of ideas comes from a Web Opportunity Workshop, or wow. Inspired by its experience with BMG, Context conducts an intense, weeklong brainstorming session in order to understand the scope and technical requirements of each project. Before consultants start writing code, they listen to a cross-section of people who will use the site that's being developed. Then they translate the project into a series of "use cases" that trace the transaction process for various types of users. Typically, a client will have a general idea of what the site should do — but it's up to Context to provide the applications and technical know-how to make the site work. Throughout the development process, a change committee votes on new ideas that could affect the direction and scope of the project. In that way, Context continues to live by a paradox that's critical to its culture — by supporting both the creative freedom necessary for brainstorming and the structure necessary to focus and filter ideas.
The ongoing communication between Context and its clients, aided by the Project Homepage, keeps surprises to a minimum. It also helps clients learn enough to manage their Web site once the development work is done. For Context, there's another benefit: fresh ideas from an outside source. After reviewing the company's test methodology, for example, one customer suggested a shortcut that people at Context hadn't thought of. As a result, Context was able to reduce the total number of steps in one of its new-product tests from 250 to 200.
Your Problem Is our Problem
It's always nice when a project unfolds as planned. It's also rare. The reality at most companies, even smart companies, is that life is filled with unexpected setbacks, problems, and technical glitches. That's why one of the big tests of the power of a knowledge network is whether it helps to solve the latest make-or-break crisis.
One of the best ways to solve an urgent problem is to find "the right people" to work on it: people who have faced it before, or who have so much expertise in a relevant area that they're likely to figure out a solution fast. At Context, those people are called "gurus." Context staffers who encounter a problem in the field simply issue a 911 message — "Contextian down!" they write — and IAN helps them find people who can help. Consultants don't have to call on anyone in particular, nor do they need to remember who knows the most about which applications. IAN automatically compares the topic of a 911 message to the technical specialties listed in the "Colleagues" records. That way, the alarm goes out only to those most qualified to put out the fire. "IAN works like a partner in this case," says Mary Durham. "It says, 'Excuse me, I know you're an expert at this. Something needs your attention.' "
Until recently, 911 messages traveled by email — which meant that gurus couldn't receive them unless they were online. Now, as well as sending email, IAN issues an electronic page. Like doctors on call, gurus are reachable anytime — which provides a valuable safety net for people in the field.
"You can feel like a lone wolf out there," says Abdou Touray, himself a guru in Net dynamics. "When you're working on a problem, you can get emotional about it. It almost takes on a personality of its own, and you think, 'I have to conquer this.' But you know there are people backing you up, and that gives you a tremendous amount of confidence. Your problem becomes their problem."
Of course, for the 911 system to work, people have to be willing to ask for help — which means that they have to admit to colleagues that they don't know something. "You do have to get over that hump," says Mike Huntziker, 38, the company's New England - practice manager. "My attitude is, I'll tell you what I don't know. Then I'll search IAN for the answer."
Some questions are more urgent than others, of course, so users designate levels of priority. "Immediate" suggests a response time of one hour or less; "urgent" suggests a four-hour response time; "routine" questions should get answered in one business day; and "not time-sensitive" questions should get answered within one week. Durham monitors response traffic and makes sure that important questions don't languish. "I believe that if a question is out there, we should make a good effort at answering it," she says.
Smart Jerks Are Still Jerks
Dror Liwer sits at a conference-room table in the heart of Wall Street, chin in hand, brow deeply furrowed. No, he's not agonizing over the company's latest e-commerce project. He's studying photographs taken during a recent office trip — whitewater rafting, along with a night of bowling — and he's trying to pick his favorite shot. "This one's great," he says. It's a group shot taken at the bowling alley, and it shows techies, project managers, and business analysts all sandwiched together, looking like a bunch of students on spring break. There's even a class clown wagging his tongue at the camera.
"He's one of our most senior people," Liwer chuckles.
It's hard to compete on ideas without technology that can store, track, and distribute what a company knows. But computers don't have ideas — people do. That's why it's impossible to compete on ideas without a group of people who are serious about creativity and who are committed to sharing what they know. In the new world of business, if you want to compete aggressively, you have to collaborate generously. That's why Context recruits according to what it calls the swan profile: It wants to hire people who are "Smart, hard-Working, Ambitious, and Nice." Most organizations embrace the first three criteria. But it's the fourth criterion that distinguishes the culture of Context from that of most companies. "I came from a Big Six firm where the one thing that all of the consultants had in common was arrogance," says Khan. "I was very attracted to the idea of swan, but I thought it was just marketing." It's not. Says Liwer: "We absolutely disqualify people if they're not a culture fit. People say, 'Are you nuts? He's got a Harvard MBA!' I don't care. If someone muddies the waters or creates an atmosphere of negativity, I'd rather not have that person around."
How do you know a nice person when you see one? Kahn offers a few of the values and attributes that he seeks out: Nice people don't just look out for themselves. They don't try to act like the smartest person in the room. They are team players. By sharing what they know, by brainstorming, and by recruiting other swans, they help keep IAN vibrant, and they fuel an overall culture of ideas. "Everybody here checks his ego at the door," Kahn says. "We have four PhDs from Ivy League schools on a project team right now, and you can't tell which ones they are."
Niceness alone is not enough. Context explicitly rewards people for sharing what they know, and it has developed ways to track which people deserve the biggest rewards. Every six weeks, for example, as part of a companywide conference call, CEO Stephen Sharp, 41, singles out the leading contributors to IAN. He receives nominations from Durham, who uses metrics that tell her who's using the network and how: She awards 1 point for asking IAN a question, for helping to answer a question, or for updating a "Colleagues" record. She awards 5 points for adding an artifact to the knowledge base. A consultant who racks up 15 or more points over a six-week period is named an IAN Good Citizen. A consultant who earns fewer than 5 points — usually a new hire — will hear about it.
Today, Context is working hard to cope with its growth and success. It invested heavily in knowledge management early in the game. Now it must focus on maintaining the swan culture it has built. But ultimately, for Context to continue to prosper, IAN must continue to evolve. "For every good idea that we capture, I wonder how many others we miss," says Ward.
In other words, IAN remains a work in progress. Which is why Durham regularly surveys users to find out how well the system is working and what needs improvement. "We don't have all the answers," Ward says, "but we're always searching for them."
That sounds like the right idea.
Chuck Salter (firstname.lastname@example.org), a Fast Company contributing editor, is based in Baltimore. We've lost track of all of his bright ideas. You can visit Context Integration on the Web (www.context.com).
Sidebar: How to Get Smarter
Selling the same old products. Failing to meet deadlines — yet again. Missing big market opportunities. These are some of the symptoms of low-IQ companies, according to Haim Mendelson and Johannes Ziegler, authors of Survival of the Smartest: Managing Information for Rapid Action and World-Class Performance (John Wiley & Sons, 1999). Unless these kinds of companies smarten up, they won't be around for much longer.
"I compare the logic of competition today to playing chess," says Ziegler, a former consultant with McKinsey & Co. and the cofounder of Synesis Management Consulting. "If you have a decent IQ, you can practice and become the champion of your local chess club. But if, all of a sudden you have to compete against a world-class player like Garry Kasparov, you won't stand a chance. The same kind of thing is happening to a lot of companies. They're basically being put out of business because they're out of their league. They don't have the organizational IQ to be competitive."
How can you gauge your company's IQ? Ziegler and Mendelson, who serves as the James Irwin Miller Professor of Information Systems at the Stanford Graduate School of Business, measure a company's intelligence in five areas.
1. Information awareness. Does each branch of your organization stay informed about customers, competitors, and industry trends?
2. Decision architecture. Is there a discrepancy between those who have information and those who have decision-making authority?
3. Internal knowledge. How often do your employees share what they know with others? How well do they learn from their mistakes?
4. Organizational focus. How fast does your company respond to new information and new opportunities? Does it deliver excellence in one or two areas — or does it deliver mediocrity in many areas?
5. Business networks. Does your company take advantage of the expertise of your partners: suppliers, distributors, and resellers?
Not feeling bright? Now comes the hard part — raising your company's IQ. Mendelson and Ziegler suggest a few exercises to get started.
1. Test your intelligence. Gather a cross-section of managers and get them to assess each department's performance in the five areas listed above. For a more in-depth analysis, do the same with people on the front lines.
2. Create a sense of urgency. Focus on a recent failure that illustrates the high cost of not being smarter.
3. Offer success stories. Describe how another company (even a competitor) benefited from increasing its IQ.
4. Don't try to get smart at everything. Start by picking one area in which you can produce major results fast.
5. Do something — even if it's small. Raising the IQ of one unit can set a standard for other units, and it can generate momentum for your entire organization.
6. Share signs of progress. If a decision gets made with unusual speed, or if critical information becomes accessible to more people than before, tell everyone.
"Increasing IQ not only improves performance — it can also make your company a better place to work," Ziegler says. "A low IQ leads to frustration." One last point: You can never be too smart. "How do you know you're smart enough?" asks Ziegler. "That's like asking, 'When is a company profitable enough?' You can always get smarter."
A version of this article appeared in the September 1999 issue of Fast Company magazine.