Who's on your short list of directors for your startup? Friends and family? The CTO? The venture firm with the heaviest exposure? And if your board meetings deteriorate into shouting matches, how will you tell the difference between valid contradictory viewpoints and counterproductive naysaying?

Entrepreneurs pulling together their first startup often try to emulate the qualities they see in large public companies' boards when assembling their own. Some traits indeed are universal: Boards should be a mix of company management, investors and interested outsiders. Seats should never exceed a dozen. And the directors should work within the general Sarbanes-Oxley rules of corporate accountability, even if an initial public offering is not planned for the foreseeable future, since established governance practices are good policy for any company, public or private.

Management board members, who can be weighted either to technology or sales/marketing specialists, mind the company's operational progress, suggesting directions that fit the vision of the founders. Investor members assess whether the operational track record meets the expectations of the venture firms and independent investors providing the funding. Good investor-directors on a board do not try to micromanage, though this principle often was ignored during the boom years, when many of the management changes initiated by boards might well have been described as coups d'etat.

Finally, a smaller but crucial percentage of the board should consist of independent members. Such directors can provide unbiased insight into whether the company's direction meshes with market trends and whether advances from unexpected directions might threaten its position.

Sometimes, very large technology companies appoint board members from well outside the industry, and startups may be tempted to follow that lead. But executives with experience on both small and large boards say neophytes sometimes adopt inappropriate rules.

Judy Estrin founded three venture-funded startups: Bridge Communications, Network Computing Devices and Packet Design. She also served as chief technology officer at Cisco Systems Inc. and sits on the boards of Disney Corp. and FedEx.

Those seats demonstrate the tendency of boards of large public companies to choose board members outside their areas of expertise. And that's fine for a company with hundreds of millions in revenue, Estrin said. But "a startup doesn't have the money or breadth to go anywhere further outside their domain than a director from their customer base — and that could lead to potential conflict problems unless that company was ruled out as a customer."

Joe Bass, founder of Monterey Networks and Covaro Networks in Austin, Texas, said he's pondered the issue but sees little value in bringing in outsiders who are any farther afield than a representative of someone in the company's supply chain. "I'd hesitate to bring someone in from the cruise industry, unless I thought we could get free cruises out of it," he quipped. "And, of course, in the aftermath of Sarbanes, that . . . would be out of the question."

Christine King, chief executive officer of AMI Semiconductor Inc. (Pocatello, Idaho), said it's possible, but certainly not easy, to hire someone close to the company's customer base without running into conflicts. AMI recently named Bill Starling of Synecor LLC to its board, reflecting AMI's high percentage of customers in the medical-instrumentation business. Syne-cor is a medical-electronics technology incubator, so Starling knows all the customer executives in that vertical market, without being a customer himself.

Estrin can speak with authority on the related issue of board "diversity," since large corporations have sought her out primarily because of her gender. A startup can keep diversity in mind when an appropriate woman or Asian-born candidate might provide a fresh perspective, she said, but knowledge of a company's technology and market is more important than establishing informal quotas. "As you get bigger, diversity in the boardroom becomes very important in setting goals and providing insights for a company that otherwise might get too settled," Estrin said. "But a board, no matter its size, should never think in terms of a diversity checklist."

King said she hopes that her recent appointment to the board of Analog Devices Inc. was more a reflection of her business experience at IBM and AMI than of her gender. "For my own board, I try to be pretty gender-blind, ethnic-blind, everything else-blind," she said. "The directors should be people of experience."

VCs' changing role
Tech-company corporate governance has gone through three phases since the venture community began to play a more critical role in the late 1980s. Before the boom years, the first-time entrepreneur with a hefty Rolodex and favors to call in could depend on the wizened sages of Menlo Park's Sand Hill Road — the locus of the VC community — and beyond. Venture capitalists like L.J. Sevin, Ben Rosen, John Doerr, Irwin Federman and Peter Sprague provided practical insight based on the old model of building a technology company with a real manufacturing base, real customers and real design talents.

During the boom years, money may have been easier to come by, but venture firms often fielded business plans and board positions out to third-string players who had little critical understanding of fast-moving technologies. What once was a disciplined investment outsider became a clueless holder of the purse strings.

"It is absolutely critical that corporate management listens carefully to directors from investment banks and venture capital firms," said the founder of one Ethernet switch company from the boom years, who asked not to be identified. But "there's a difference between a knowledgeable tech investor and someone from the [Jack] Grubman and [Henry] Blodgett school," the founder added, referring respectively to the Salomon Smith Barney and Merrill Lynch analysts who infamously championed weak stocks at the start of the telecom free-fall.

"When you have a board member coming in from the automotive or plastics or consumer goods industries, they usually are careful listeners and don't pretend to know much," the source said. "But the venture guys from the late 1990s really didn't have much of a clue and liked to pretend they did. Their demands were based on herd behavior rather than solid analysis of technology and markets."

Another source, who has worked for three different Sand Hill Road firms, said startups are justified in their lingering fears that not all superficial partners have been purged from venture firms. The venture community is tighter and more restrained than it was in the late 1990s, this junior partner claimed, but there remains a mix of senior and junior partners who are less knowledgeable about market conditions and who retain a boom-market mentality of raid profits.

The tech recession and subsequent toll of Sarbanes-Oxley legislation have led to a third and very sober phase of corporate governance. A new crop of principals at venture firms comes from the generation of 1990s-funded companies that survived the crash. Consequently, startups can pick from investment executives with real battle scars and a sober sense of what might actually work in the 21st century.

But prospects are tougher, and not just because venture money is harder to come by. The Sarbanes-Oxley rules, in a bid to address the public's concerns about board accountability, has placed new demands on directors to be more active board members. Thus a venture capitalist who might have served on five or six boards in the 1990s would be hard-pressed to serve on two or three today.

"Is it harder to get board members today? You bet!" Estrin said. "And the real problem . . . is that you just can't afford having a director who neglects to do homework. Some venture partners still are on way too many boards, and you must have a board member willing to do homework in between meetings. Coming into board meetings unprepared simply leads to needless discussions that waste critical time."

Bass said the better venture firms are realizing they can't manage 10 or 12 companies effectively and are limiting themselves to an informal maximum of 1.5 to two deals annually per partner. Sometimes, a junior partner within a firm that understands a vertical market is the best candidate to serve on a board; elsewhere, the senior partners have the better insight.

A startup with five venture investors might hope to fill its three open board seats with the best partner candidates from among the five VCs, but rarely is the hope realized. That's partly because venture firms can designate their own board members for companies in which they invest.

Thus, Estrin said, it is often wise to choose investment companies based on the people they might put on your board. In the 1997-2000 period, she said, investment firms were chosen largely on their ability to create buzz, but hunting for hype can be hazardous in the current environment.

Bass agrees. "Pedigree mattered more in the boom years than expertise, and as a result, VC partners got put on boards with detrimental results in some cases," he said. "Fast-forward to the post-bubble years, and the world has shifted. You have to look at the firms funding you on the basis of the expertise they could bring to your board."

After a company has gone public, the venture investors often no longer sit on boards, though some may voluntarily remain because of personal interest. It is important that a public company keep the internal CFO involved and that it have outside board members with financial expertise. King of AMI said that Colin Slate, chief financial officer of Tektronix, and Atiq Raza, founder of Raza Microelectronics and former president of Advanced Micro Devices, are board members of AMI primarily because of their financial savvy.

Diversity on a board can mean more than its ethnic and gender breadth. It is critical for a startup to pick board members whose expertise complements that of the management members of the board. That means finding marketing and finance experts for a technology-heavy management team, for example, or finding technology-savvy outsiders if the founders are skewed toward marketing expertise.

It also means finding the right mix of personalities to allow boards to hash out differences without a meltdown. Some 1980s-era companies, notably Cypress Semiconductor and Apple Computer, were known for their hostility-ridden meetings. In the late 1990s, the board meetings of some Internet startups moved from screaming matches to fisticuffs.

While the tenor of board meetings has now swung back toward civility, it's not necessarily wise to shoot for a consistently mellow tone, particularly if that means rubber-stamped business plans.

Calm vs. fiery
"The personality of the board as a whole depends a lot on the chairman and CEO," Estrin said. "In most cases, mild-mannered may be better, but not if the management team is locked into a single way of thinking as a result. If the corporate management has a tunnel vision and is not on the lookout for competitive ideas coming out of left field, then it may be better to have board members ready to stir things up. If management is innovative by nature, then outside board members should be a calming influence."

Bass noted that board members are difficult to remove if they become a source of needless problems or challenges to the corporate vision. A person must have confidence to play devil's advocate, Bass said, which can be a good thing for a board. But some members, from the investment community and outside companies, cross the line from self-confidence to arrogance and question everything the company does.

"It's always good to have outside directors with a different point of view, because rocking the boat can be important," King said. "But there comes a time when a board should be able to reach consensus and carry through on the basis of that consensus."

The simple reality for many startups, executives and investors agree, is that beggars can't always be choosers. Not only is money tight today, but the time constraints of outside directors, amplified by the accountability requirements of Sarbanes-Oxley, mean corporate founders have limited freedom in assembling a board.

In such an environment, the first rule is that the requirements are too critical to fill your board seats with casual friends, angel investors with little industry knowledge or overextended third parties.