Driven by ever-increasing equity market valuations and growth challenges, many large corporations are bestirring themselves to build new enterprises that have the potential to grow much faster than their core businesses. Yet in doing so, senior managers at the parent companies often make the same well-intentioned mistake that children do with their first goldfish: overfeeding.

It seems to be common sense that if you seriously want to build a business quickly, you shouldn’t skimp on funding and other resources. Many senior managers believe that the financial and technical support the parent corporation gives to its start-ups is their main source of competitive advantage.

Although this is often the case when companies invest in projects to extend core businesses in familiar or strongly related markets, when they invest in new businesses the opposite may well be true. Apple’s Newton personal digital assistant (PDA), General Electric’s "factory of the future," and a host of other unsuccessful new ventures show that lavishing capital does not guarantee success.

In fact, bestowing too many resources on new ventures undermines the discipline they need to grow. Excessive amounts of capital can make the managers of a new venture expand its product range too quickly, invest in too much infrastructure, and delay going to market for too long—all potentially fatal mistakes. By contrast, successful start-ups, such as Excite (Internet media), Amgen (biotechnology), and Tivoli (network management software), grew and prospered under the constraints of venture capital-style funding.

The logic of these constraints, which force new ventures to focus and to limit the degree of risk they impose on their sources of funding, rests in part on the fact that few such fledgling enterprises turn out to be successful. For corporations, the challenge is to ration the financial resources they provide in the early stages and to concentrate on exploiting their nonfinancial resources, such as access to people and partners.

The power of penury

A large technology company recently attempted to build a business in the ill-defined but potentially very large area of home automation. As enthusiasm for the project swelled, this effort to bring a single product to market ballooned into a venture aimed at developing a suite of six new products, each designed to provide a different type of functionality and targeted at a different market segment. Instead of creating a winner in a single niche as a platform for building a broader position, the company spawned a full line of mediocre losers that were shut down within two years.

By contrast, consider the success of independent start-ups funded by limited amounts of venture capital—projects that focused on winning in a niche before attacking related markets. A highly visible example is Amazon.com, which reached a market capitalization of more than $30 billion in 1999 by establishing itself as an on-line retailer in a number of segments and as a provider of related services, such as on-line auctions. Amazon reached those heights by starting with a highly focused model in the niche of on-line book sales. It was then able to refine its initial approach to on-line retailing quickly by intensively gathering data on the usage patterns and buying behavior of its customers. With this understanding of the marketplace, an enticing brand, and significant market credibility, the company has been able to launch itself successfully into other kinds of retailing. Indeed, its most significant advantage as it takes on more experienced and established retailers of all shapes and sizes may be the insights about customer behavior it learned in the early days.

The experience of a product development leader at a successful biotech start-up shows how a tight-fisted approach to funding sharpens a new venture’s focus:

We had to push our device over one major technical hurdle before we could do the IPO. I estimated it would take $8 million. When I was told we had $2 million and no more, I thought we were finished. But we had no choice. In the end, we did it for $2 million, which required all of our engineers to drop work on several new concepts. Looking back, if we had had the $8 million I might never have gotten our engineers to drop their pet projects to ensure the success of our principal product. The IPO would have never happened.

Too much capital also encourages businesses to build up facilities and staff needlessly. GE’s failed factory-of-the-future venture started out with $70 million in funding. As might be expected from a company with a traditional corporate mindset, management’s first step was to build a $35 million headquarters building and fill it with employees. Despite the high level of resources available, the new business failed to develop a compelling product for its customers. After four years and several hundred million dollars, it was scrapped. Providing such a lavish infrastructure not only is wasteful; it also creates a degree of comfort inimical to the sense of urgency every successful start-up must generate.

Moreover, very early in the life of a new venture, it must attract talent and establish relationships with customers and potential partners. In fact, an additional pitfall of excessive funding is the way it encourages the managers of new ventures to do everything in-house, even if better and more economical suppliers and solutions lie outside. Persuading a supplier (or an investor, a customer, a potential partner, or a talented new hire) to assume some of the risks of a new venture is itself a form of market validation.

Large corporations starting new businesses, however, sometimes take a different approach: they throw their big, number-crunching marketing staffs into the fray. Dependence on detailed market analysis may be acceptable in well-developed businesses, but it is a recipe for disaster in the world of new ventures, where businesses must evolve rapidly as they interact with a dynamic community of customers and partners. Internally oriented market analysis is a poor substitute for early validation in the marketplace—a more telling indicator of a product’s longer-term prospects—and the adjustments that help make the product successful.

The contrasting development of the Newton and the Palm Pilot shows the importance of active market evolution

The importance of active market evolution is shown by the contrast between the development of Apple’s Newton and of Palm Computing’s Palm Pilot. The Newton was conceived as a palm-sized alternative to large desktop personal computers dependent on keyboards. Apple hoped that a wireless portable device with the ability to interpret handwriting would overcome resistance to PCs. But after devoting ten years and more than $500 million to the Newton, the company shut it down in 1998. Not coincidentally, Apple’s six-year initial development of the Newton was an almost entirely internal effort without direct exposure to market feedback. In the end, customers did not accept the Newton, which was too expensive and big and also lacked the functionality desired by specific customer segments.

A very different strategy guided the development of the Palm Pilot, whose success was ultimately as great as the Newton’s failure. The Pilot’s creator, Jeff Hawkins, a former Intel engineer and an expert on human memory and cognition, left his programming job at GRiD Systems in 1992 with a paltry $1 million, which he obtained from Tandy (GRiD’s parent) and two venture capital firms. After 18 months and an additional $2 million, Palm Computing launched its first product: the Zoomer PDA, marketed under the Tandy brand name. By then, Palm had built relationships with Casio to develop the hardware, with Geoworks to create the operating system, and with Intuit and America Online to provide more software. Palm then used its ties to Casio and Tandy to secure distribution channels.

The product sold at first for about $700, like the Newton, and offered roughly the same functionality. Like the Newton, too, Palm’s first product failed, and for similar reasons. When the Newton ran aground, Apple tried to reposition the product by making small technical modifications. Palm, by contrast, received market feedback that revealed a need for cheaper devices complementing rather than replacing desktop PCs. By offering only a calendar, an address book, and note-taking facilities, coupled with a simplified "Graffiti" language for user input, Palm cut the cost of the device by more than half, to $300. When Palm’s original manufacturing partners showed little interest in the new version, the company found contract manufacturers to replace them. These changes consumed an additional $5 million in venture funding and in stock options for employees and business partners. Today, the family of Palm devices generated by the original Palm Pilot dominates the market for hand-held organizers. Three million units have been sold in the first three years.

As the stories of the Newton and the Palm Pilot show, success in new ventures often requires an approach radically different from the more traditional one used to expand existing businesses incrementally. The CEO of a successful Silicon Valley software start-up observes that the original concept of most companies like his represents "only the platform for the team to figure out what it is really going to do." As a result, successful independent start-ups rarely implement their strategies as originally conceived; the strategy evolves with the business.

Milestones

So the key question is this: how can managers whose instincts have helped them make successful decisions in core businesses reprogram themselves to think and act differently in making decisions about more unconventional ideas for new businesses?

Some corporate managers involve themselves in technical and day-to-day tactical decisions to keep the new venture—in their terms—on track. They insist on judging its short-term financial performance as though it were part of the core business. Often, this kind of behavior is all that is needed to drive entrepreneurs out of the company and to kill new undertakings before they have any chance of success. Contrary to what managers’ instincts tell them, revenues and earnings are not a good measure of a new venture’s performance in the early stages. The few start-ups that survive typically go public after four to six years.¹ They might then be valued at $100 million to $200 million, though most have minimal or no earnings at this point. If such companies as Cisco Systems, Genentech, and Yahoo! had been evaluated in their infancies on the basis of the near-term earnings metrics of large corporations, they could never have become the business legends they are today.

In the absence of earnings, the challenge is to find some other way of measuring a start-up’s value in the market. One approach is to execute a partial spin-off.² The sale by Hughes Electronics (then a satellite manufacturer and defense electronics supplier) of 2.5 percent of DIRECTV to AT&T caused its share price to jump by more than 20 percent. The sale price implied that DIRECTV had a market value of $5.5 billion, despite its total lack of earnings to date. This was only one of a series of market-validating partnerships Hughes built over the course of the system’s development. The others included ties to Thomson in manufacturing, Matrixx Marketing (now merged into Convergys) in customer service, and DEC in billing systems.

But even partial spin-offs are worthwhile only after a new business has evolved significantly. Perhaps the key challenge facing corporate managers is how to make financing decisions about new ventures at earlier stages of their development. The answer is to create and enforce predetermined milestones, such as beta tests, the number and quality of partnerships and customer relationships a venture manages to create, and success at external hiring. This is the framework that "outside" venture investors rely on to make decisions about funding. Such milestones are rational expectations that keep the participants in new ventures motivated and help prevent their overseers from wasting money. More funds are released only when a milestone has been achieved; otherwise, they are withheld no matter how much time has elapsed since the previous milestone was reached. Exhibit 1 shows how this approach might work, though the timing and levels of funding depend on the industry.

Tivoli’s network management software business, for example, grew in a series of stages by achieving milestones based on external relationships and market tests rather than financial performance metrics (Exhibit 2). Founded by several IBM employees who could not obtain acceptable support from the company to pursue their ideas, Tivoli was ultimately acquired by IBM for $743 million after developing revenues and earnings and has prospered ever since.

Corporate new ventures that abide by the milestone framework—whether they remain internal or are eventually spun off—gain transparency, which allows the market to assess their value at every stage of development. But while transparency is certainly preferable to confusion, ignorance, and surprise, for investors it can do no more than expose the degree of risk involved in new ventures. And particularly in the early stages, the risk is high. This has two effects on seasoned venture capitalists: it inspires them, first, to demand returns (often 60 to 80 percent a year) reflecting that risk and, second, to restrict the amount of funds they commit to any particular venture at any given moment (Exhibit 3). Only as the risk declines over time do they release new funds.

Any money withheld can be funneled to a host of other new ventures, since the odds that a particular project will fail are substantial, but one that succeeds against all odds can compensate its backers for a multitude that do not. In short, corporations, along with their venture capital brethren, must realize that building new businesses is a portfolio game and that stinting on money is just as important as providing it.

Right idea, wrong team

Although getting the right level of focus and insisting on higher and higher levels of market validation are key aspects of the milestone-based process, the management team’s quality and capacity are equally important. Over time, investors in start-ups have learned—sometimes the hard way—that picking the right team is the ticket to making portfolios successful. As Arthur Rock, widely regarded as the father of venture capitalism, puts it, "If you find good people, they can always change the product. Nearly every mistake I have made has been picking the wrong people, not the wrong idea." Venture capitalists have learned that if they focus a really bright and experienced team on a market space capable of yielding results, the team, with the right incentives, will find a way to make the venture work.

Even if companies recognize that a strong leadership team is essential for building a new venture, they typically fail to understand the types of skills it requires. This difficulty increases when the scale of resources targeted at the new venture increases its visibility. The manager of a new venture formed by a large biotechnology firm described the problem:

Now that we are receiving substantial corporate investment, the heads of each of the divisions want to see a management team with a background similar to theirs. What they don’t realize is that, given the pace of this new market and the scarcity of resources, their management teams could never run our business.

Nonetheless, the idea that only managers who have outside entrepreneurial experience are qualified to lead new ventures is a myth. It is true that people with histories of strong performance in a company’s established lines of business are accustomed to substantial support from staff and systems and such managers’ performance often suffers when they are forced to work with limited resources in a rapidly changing environment. But most CEOs of successful independent start-ups have backgrounds in large corporations, though the nature of their corporate experience typically differs from that of their fellow managers.

When an executive search firm sets out to recruit senior people for start-ups, it often looks for a record of building, if not a whole new business, at least an office in a new location or a new brand. Managers with this kind of background already understand how to work with limited support outside the framework of quarterly earnings objectives. A large corporation has many such men and women, though its senior management may not realize this. Indeed, the top managers of a leading electronics company once attributed its poor record in building new businesses to a dearth of internal entrepreneurs, but venture capital and executive search firms consider that very same company to be a leading source of start-up talent.

Leadership teams for new ventures must be established early (Exhibit 4) and include a wide range of skills. Although new corporate ventures are rarely short of technical talent, big companies usually don’t see to it that new businesses have the kind of marketing, financial, and sales capabilities that venture capitalists demand in the start-ups they finance. Leadership from outside the corporation—innovative people who have links to customers and potential partners—enriches a new venture’s management team. At Xerox New Enterprises, a division that commercializes novel technologies, the CEO of every new company comes from outside Xerox.

To attract external and internal talent, corporations founding new ventures will likely have to modify their traditional approaches to compensation. In established businesses, strong individual performance is typically rewarded by increased compensation in the form of bonuses or stock options. Measuring the performance of individuals is essential because the relationship between their actions and the performance of the corporation as a whole is often weak. But in a start-up with constrained resources, the performance of leading individuals is highly correlated with the performance of their companies, so linking rewards to share price appreciation is perfectly reasonable. As an important by-product, this approach encourages collaborative behavior: if the venture fails, everyone loses. Although it is harder to link compensation to the collective performance of a nonpublic corporate new venture, it is possible to do so by basing salary increases on the attainment of milestones or, even better, by giving managers a piece of the business.

The rewards individuals earn in a new venture should reflect the risks they face, not their corporate positions

It is important as well that the rewards individuals can earn in new ventures reflect the risks those people face, not their positions in the corporate hierarchy. As one venture capitalist recently told us, "Relative to corporations, we have a huge advantage in attracting talent: our funding processes give venture managers a higher probability of success, and our compensation systems create a significantly higher level of compensation for them when they succeed." While this will mean that the managers of a new venture, at least for a time, earn less than their counterparts in the core business, they will be in a position to win large bonuses or grants of options once the new venture makes good. Such an incentive structure also motivates managers to make a "once-in-a-lifetime" effort that, if successful, will create lifetime financial security for them and their families.

A constrained approach to funding strengthens these types of incentives, since the largest part of a manager’s compensation is deferred until the worth of the new venture becomes clear. This also has the benefit of keeping all but the most serious entrepreneurs from joining internal ventures. Xerox uses another device as well: managers who move to one of the ventures in Xerox New Enterprises must give up the standard benefits package. This is enough to discourage many people who lack entrepreneurial instincts from joining.

Achieving success

The natural tendencies of many corporate managers inadvertently thwart their new ventures because they focus on the relationship between the corporation and the venture instead of thinking about it in a broader context. Exhibit 5 represents this context as a circle of interactions among the venture’s owner, the venture itself, and the innovative community in which it must succeed, including all of the relevant leaders in a new venture’s market—customers and potential partners, for example. Although in some industries, these communities may be concentrated geographically, as they are in Silicon Valley, in others they spread across many areas.

Managers can improve their prospects for success in a new venture by ensuring that its linkages to the innovative community are as strong as its linkages to the parent corporation. Unfortunately, few big companies evaluate a venture on the basis of how good it has been at attracting partners and external talent, for example, early in the venture’s life. They instead focus on the more familiar ground of evaluating financial performance and allocating resources.

Corporate managers must not succumb to the temptation to throw money and resources at new businesses. Conventional wisdom may tell us that for want of a nail the kingdom was lost, but when it comes to financing a new venture, parent companies are far better off letting its managers figure out a way of doing without the nail.