Not long before the Christmas season began, I received a vaguely anguished e-mail from a former student of mine named Jane, who had just taken a job with a New York e-commerce consulting firm. The company had everything it needed, she said: more clients than it could handle (including several "bricks-and-mortar" retailers trying to click into the Internet Age), steady revenue, a good dose of venture capital, a fast-growing staff of powerhouse developers, and a pair of hands-on, charismatic founders. But nobody there seemed to know how to get the work done.
Although she had joined only a few months before, Jane was already a seasoned veteran, the "old-timer" that employees went to when they needed to settle disputes. And there were many disputes. "Last week, two of our people got in a snit," she wrote, "with clients copied for the entire e-mail exchange. The two of them sit next to each other but they rely on e-mail because face-to-face takes too much time." Everyone was working 65 hours a week ("80 is not unheard-of," she said), but they were spending much of that time undoing one another's mistakes after the fact. The founders said they wanted to make every team autonomous, but they insisted on approving all major inquiries to clients. Jane had started to write up standards for the company's work, a way to track whether they were fulfilling their clients' demands, but she got pulled to another project - a billable project - before she could get far. Trying to deal with all of this, she'd browsed the "Fast Company" Web site and read everything she could find elsewhere about organizational learning. "But it's all aimed at changing older companies. I'm in a startup. I feel like I'm digging for a needle in a haystack. And I'm not sure where to go next."
I'm withholding Jane's real name, the name of the company, and the exact nature of its business, but the rest of the details are scrupulously accurate. And complaints like this are popping up more and more these days in New York's Flatiron District, in the Internet businesses of Washington D.C.'s beltway, and even in Silicon Valley. If you have watched the phenomenal market capitalization growth of Internet companies with envy and alarm, then you can take some comfort, perhaps, in stories like Jane's. These companies are carrying the seeds of their own destruction, and they will probably inflict worse damage on the old-line manufacturing and retail companies that enter into partnerships with them.
"The dirty little secret of the Internet boom," says Christopher Meyer, who is the author of "Relentless Growth," the 1997 management-based-on-Silicon-Valley-principles book, "is that neither startup wizards nor the venture capitalists who fund them know very much about managing in the trenches."
In the last few months in particular, there's been a growing sense that something is wrong. Dot-com company executives are beginning to try the same kinds of "employee-empowering" measures that their mainstream, Fortune 500 counterparts have tried: Talking about values, building shared vision, even hiring facilitators to run meetings. Six months ago most of these ideas would have been dismissed out of hand as hopelessly old-fashioned. But amid the frenzy of the Christmas 1999 e-commerce season, the management fashion shifted. "I have a client in his fifth startup who's decided that he's going to do it differently this time," Silicon Valley management consultant Bill Underwood said in November. "He's going to 'create a culture.' This is the fourth person who's told me something like that."
Anyone who has tried to create a culture, however, knows it can't be done on Internet time. Cultures aren't designed. They simmer; they fester; they brew continually, evolving their particular temperament as people learn what kind of behavior works or doesn't work in the particular company. The most critical factor in building a culture is the behavior of corporate leaders, who set examples for everyone else (by what they do, not what they say). From this perspective, the core problem faced by most e-commerce companies is not a lack of culture; it's too much culture. They already have two significant cultures at play - one of hype and one of craft.
These cultures of hype and craft, of course, aren't limited to Internet and e-commerce businesses. They've existed in mainstream manufacturing and retail corporations from the beginning of corporate history; indeed, they were prominent in the guilds and joint stock corporations from which the corporate form developed. Hype and craft are generally the two strongest cultures in a company's early life, as manifested in the cultures of the founder/entrepreneur and the first R&D/engineering/production team, respectively. But during most of the 20th century, as companies matured into mainstream corporations, other cultures - those of finance, labor relations, marketing and managerial bureaucracy - eclipsed and overwhelmed the cultures of hype and craft.
Only in a few industries have hype and craft remained prominent as corporate cultures. Whether by coincidence or for some fundamental reason, these have included the most critical industries of the Internet Age: Personal computers, Internet technology (but not mainstream telecommunications), software and media content. People in most of these industries have learned to maintain a delicate balance between hype and craft. But under the high-speed pressure to grab Internet market share, the balance is falling off kilter, and many e-commerce companies are faltering as a result.
Consider, then, the culture of hype. This represents the visible ethos of venture-capital-driven Silicon Valley, where 25-year-old C.E.O.-wannabes drift from V.C. to V.C. pitching their business plans, living out of their cars, unable to pay their cellular-phone bills one week, and granted millions in stock options the next.
If the corporate culture of many Internet and e-commerce startups feels like a fraternity ethos, with brash, arrogant claims and tests of manhood built into every deadline, that's because the founders have often come from a background of Internet strategic planning and promotion. Their income, and their success, depends on the number and prominence of deals they can cut; being able to say that they've taken on a job for a prominent client - say, eBay or Brooks Brothers - gives them the cachet to raise more money. Naturally, they are easily drawn beyond their original scope, jumping from developing 200-page Web sites to developing mega-page portals with database engines that compose an unlimited number of pages dynamically. After all, scale is where the money is: A salesperson taking on the conversion of the entire global retail infrastructure can't afford to be shy.
Most Internet-related startups and divisions are run by people of hype, familiar with the methods of raising capital, gathering people together, and articulating new ventures. They are propelled to "Internet speed" not just by their love of action and contact, but by the immense costs of startup growth - the need to keep expanding rapidly to dominate the market and fend off equally voracious competitors. In e-commerce, these costs are exacerbated by the unexpectedly high costs of the e-commerce transaction engine - the database-driven software that displays merchandise, tracks and fulfills orders, actively calls up pages from a server, and coordinates with all the partners and intermediaries involved in each transaction. Because these technologies are so new, there are no generally accepted, standard building blocks for e-commerce software. Thus, a surprisingly large number of firms feel forced to build their own or adapt existing packages at enormous cost. "They think a commerce engine will cost half a million," points out entrepreneur Ken Ketch, founder or board member of half a dozen Internet startups. "But it ends up costing several million. Then they have to justify those costs in terms of the stock price."
Meanwhile, the work has to be delivered and promises kept. Thus, the second culture, the culture of craft, comes in. Programmers, designers and practitioners of the new profession of information architect are all artisans at their core. Like members of a craft guild, they like to delve deeply into a project, come to an understanding, and deliver an elegant solution. Even when individual artisans are sympathetic to the hype ethic (or when they stand to make millions from it), the culture of craft is innately persnickety, recalcitrant and suspicious. (It has to be, because craft work requires moving into a hype-free, reflective mental space where there is nothing but the hum of the work, where effort takes place on a semiconscious level, where the writing and design flow through the mind in a way that will not suddenly shift gears just because a new, more insistent client has arrived.) An artisan/developer composes, note by note, the structure of a Web site in a way that no one else can. This frame of mind is antithetical to, say, the mindset needed to convince a V.C. to invest in a company. The two frames of mind cannot exist in the same person at one time.
The hype people, if they're smart, learn to protect the craft people, instead of draining them. They know that sophisticated craft people need clear boundaries within which to make choices, time for reflection, and a chance to express their ideas. Similarly, smart craft people gradually learn, sometime during their career, to value the strategic primacy, and the groundbreaking audacity, of the culture of hype. They even learn to speak effectively to the strategists. But HTML and Java wizards in their early 20's typically haven't learned how to do that yet, and they won't learn it while under the gun of a Christmas Web site deadline.
The result is debilitating and thoroughly unnecessary culture clashes. The management style that brings hype people success in their world makes it difficult for them to manage craft people effectively. I know a young San Francisco-based Internet producer/designer (we'll call him Jim) put in charge of a client Web site after a few months in the field. He's desperately craving permission to tell his clients: "You can't have every feature you want in this site. Not in the time frame we have. Not at your budget." But Jim's boss, a skilled courtier of investment sources and customers, has learned to block messages like that. And Jim doesn't have the experience to know how to contradict his boss's demands for the good of the project.
Because the need for help is so great, and the deadlines so urgent, craft people are commanding unprecedented amounts of money. It's almost a cliché that, whatever you offer a skilled Web site architect right now, someone else will double it within a month. Yet craft people are not motivated by money in the same way that hype people are. By and large, they want to be comfortable, and they are delighted by the idea of stock options; stock options signify that the organization values them. Most of all, however, they want the chance to see what they can do, on a team or off it. They don't want decisions to be driven by hype; they want to know that they can feel proud of their work. "They are motivated to build something elegant," notes Silicon Valley organizational-development consultant Lesly Higgins. "They are not motivated to meet a deadline that usually requires a compromise of elegance."
Hype people will agree with "elegance" in principle - who wouldn't? But when they get caught up in the speed of raising cash, they don't want to hear about it. Instead, they take on new business. "We need the uplift," they say. Thus, the craft people default to the only form of reward they have left: demanding higher salaries or better stock options, which in turn adds to the urgency that the hype people feel, which puts more pressure on the craft, and so on, until the e-commerce industry spins out in a frantic whirl of mutual misunderstanding.
But it doesn't have to be that way. Instead of self-consciously trying to build a corporate culture, or aping the worst bureaucratic tendencies of the dinosaur companies they are replacing, there are three things that Internet and e-commerce startups - and old-line companies entering these new businesses - can consider to sustain themselves past, say, this summer. These ideas are tough to implement at breakneck speeds, unless you are willing to think through your work design right at the beginning. And they're counterintuitive, to some extent, to both the culture of hype and the culture of craft, because they start with an explicit appreciation of the differences between the two cultures - differences that are generally unobserved.
1. CULTIVATE TEAM CULTURE
At the e-commerce consulting company where Jane (the young information architect who e-mailed me) works, a team consists of about 20 people: 12 or 15 programmers, a project manager, and five or six designers. They work consistently for three or four clients, dividing their time among the clients' offices. They are supposed to work in synch but often forget to make time to come together. And some of the clients' staff work with them so intimately that, in effect, the client people join the consulting team, rather than the other way around. The net effect is that, while the culture of hype is operating at a company-wide, strategic level, the work of craft takes place at the level of the team. If Jane's company falls apart, her team could very easily pick itself up and move to another company, nearly intact. And that would make it much more effective than simply creating another team from scratch.
It is currently fashionable to say that the old, tightly knit mentoring relationships of bricks-and-mortar companies are dead, that individuals are now responsible for their own development and career growth. Unfortunately, this view is not sustainable; there are too many risks, even in a high-growth economy, and too much human waste. The task of developing people will move away from companies, since they are not stable enough; it will move to the team level. In other words, if success depends on building a new "culture," that effort will have a lot more effect at the team level than on any company-wide level. It's reasonable to expect, in the turbulent e-commerce business environment, that companies won't necessarily evolve intact cultures. But teams do; as one e-commerce veteran puts it, they're "islands of stability in a place where nothing else is stable."
In Charles Ferguson's brilliant memoir "High Stakes, No Prisoners," he describes how the team that produced Vermeer's FrontPage stayed together after Microsoft bought the company, remaining intact and even a bit countercultural within the larger company, kept together as long as they produced results. Mr. Ferguson attributes (as do others) Microsoft's success, in part, to its reliance on this structure. Similarly, Apple is less a single corporate culture than a culture of singular teams, a key factor that has sustained the company through its many travails. As Ferguson points out, this structuring mimics the modular structure of software architecture, but it is also the way that most Fortune 500 companies achieve their innovations. Below the radar of the corporate culture, with or without executive approval, teams accomplish miracles. Sometimes these teams exist for years, moving from project to project and growing continually more capable as team members internalize one another's thinking.
But many e-commerce teams lack what they need most: a clear chain of accountability, with a leader who can both manage the team and choose its members. This leader needs to understand both the craft and the hype cultures, since the team will have members from both and report to both. The team should have enough deliberate group process practice, even amid the melee of deadlines, for the diverse hype and craft people to learn to think in alignment. Teams should regularly conduct post-mortems, for instance, meeting every couple of weeks to look over their past work and reflect on changes for the future.
Only teams that stay together, through various projects and even various corporate parents, can build their collective capabilities: their seamless complement of skills and personality. Only intact teams can work their way through the awkward phases of collaborative decision-making to the point where they know how to think together, and their work has sped up dramatically. Teams that explicitly foster that development will become much more attractive to people, particularly craft people, than teams that simply treat people as commodities. People will gravitate toward the teams, not the companies, that they really want to work for. These teams will become much more attractive to clients as well, because they will produce results more easily.
2. CREATE AN INFRASTRUCTURE THAT SUPPORTS INTER-TEAM LEARNING
Perhaps the most critical job for Internet startup founders, then, is not raising the money but setting up a context for the teams that produce - without being dominated by dysfunctional internal politics. They need to know when to get out of the teams' way, but also when to get in their way - with organizational infrastructure that makes it natural and easy for teams to exchange knowledge, lend each other key people as needed, and learn from each other. A very few Internet companies, such as the Viant Corporation, have taken this lesson to heart.
Creating an infrastructure might mean setting up rotating assignments in which people leave a team, move to a company-wide support group like specification-writing or customer service, and then return to the original team. It might mean paying a lot of attention (as Viant does) to electronic communication among teams, or to co-location: changing the location of people's desks as teams evolve, so they can collaborate more informally with critical colleagues. It might mean an elaborate set of cross-team symposia, at which senior managers regularly show up (in the same way that General Electric Company chairman and C.E.O. Jack F. Welch Jr. established the "glue" of G.E. by showing up regularly for sessions at G.E.'s Crotonville, New York training center). It does not mean creating an elaborate staff support structure for human resources and other common functions; these should be handled by the teams themselves, and as much as possible by line managers, or they will become self-sealing bureaucracies.
Here is one area where a seeming waste of time - bringing together managers from a group of teams to establish standards for the company's projects, for instance - can save much more time than it requires. Teams in many companies don't even realize when they are reinventing each other's wheels, because there is no one to tell them. Often, customers, headhunters and competitors know before the protagonists do. The only way around this is to set time aside to compare notes, with the founders in the loop as much as possible.
People from the hype culture often miss this point. They see knowledge as something to be hoarded and held for advantage, even within their own companies - an ironic stance for an industry founded on the idea that "information wants to be free." But founders who hide their inside knowledge and perspective from their people are sure to short-circuit the company's capability. Knowledge-sharing software, no matter how well-designed (or expensive), can only go so far, unless there is a cultural predisposition to spend time in mutual inquiries. This means people at the top of the company have to take part in those online exchanges. (Microsoft Corporation chairman and C.E.O. Bill Gates, for instance, is known for his attentiveness to internal e-mail.) Craft people, in particular, have a great deal to learn from hype experience, and they will repay that knowledge with much more effective work, and with a kind of loyalty that high-tech companies cannot otherwise engender.
3. PROMOTE EXPERIMENTATION AT THE TEAM LEVEL
Ultimately, I suggested to Jane, all the organizational-learning techniques in the world wouldn't do her any good unless she were willing to go to her bosses, the startup's founders, and say something like this:
"If you let me build my own team, and choose and develop the people, I'm willing to take on [name of tough, challenging project here]. But I want to take our own development seriously. I want to try some new ways of organizing the work, regularly evaluate them, and try to learn how to manage ourselves in this new territory. After a few months, we'll come back together and see what we've accomplished, and which of those innovations might apply to the other teams around here. But it will only work if you give our team enough autonomy to learn from our experiments."
The founders should then talk openly about their criteria for success: What would make them think that a project was working well? (If they have no clear criteria - and a startling number of hype people have never thought about this on the craft level - then there's all the more reason to let these experimental teams flourish.) And every six months or so, there is value in a short retreat where the experimenters can meet to consider their innovations - and their mistakes.
This means giving craft people responsibility for managing production (and a lot more autonomy in client relations.) And it means establishing a precedent where craft people have reason to feel committed to the company for the long haul, not just because they'll get stock options, but because they'll be treated as knowledgeable professionals. The alternative is a culture where craft people feel continually tempted by higher bids from competitors, and where hype people have to focus even more on raising money for the next wave of expansion.
All three of these measures start from the assumption that change in high-tech companies will come from the bottom up (from craft to hype) as much as from the top down (from hype to craft). The bottom, after all, is not that far from the top. If the hype people (who are often at the top because they're closer to the financing) think of the craft people merely as "resources" to be hired and deployed, then the entire business will be akin to a commodity. If the hype and craft people recognize what they have to learn from each other, and how much their businesses depend upon both of them, then they can engender a culture greater than the sum of its parts, and perhaps a culture resilient enough to withstand the turbulence of the next few years. That culture will probably be team-based, because teams will have so much more to offer people than companies will.
Reprint No. 00103
firstname.lastname@example.org is the “Culture & Change” columnist and a regular contributor of “The Creative Mind” profiles for strategy+business. He teaches at New York University’s Interactive Telecommunications Program. His Web site is www.well.com/user/art. Mr. Kleiner is the author of The Age of Heretics (Doubleday, 1996); his next book, Who Really Matters: The Core Group Theory of Power, Privilege, and Business Success, will be published by Doubleday Currency in August 2003.