Originally published in Red Herring issue of November 12, 2000.

Every decade or two, corporate chiefs and consultants get it into their heads to change the way companies are designed. In the ’70s, it was fashionable to be a diversified conglomerate, with interests in cookies, oil wells, and everything in between. The ’80s were all about taking those conglomerates apart. (Time to “focus on our core business,” companies kept saying, in press release after press release.)

But now, just like platform shoes and ’50s hairdos, the concept of big is making a comeback. It started earlier this year with incubators, companies that claimed they could sprout other companies, side by side in a single office building, like so many hothouse tomatoes. Together, the offspring were supposed to form a new kind of conglomerate, of the add-venture-capital-and-stir variety. Working in close quarters, the new companies were supposed to thrive on synergies, which were often promised but rarely manifested. A recent Harvard Business School study, “The State of Incubator MarketSpace,” counted more than 350 incubators.

But the law of gravity has caught up with this fad, dragging down its biggest, most vaunted examples, from CMGI to Divine Interventures, which bodes poorly for the planned public offering of Idealab, the company that started us all on incubator madness.

If the incubator is a failure, what’s next? Anil K. Gupta, a visiting faculty member in the Stanford (University) Technology Ventures Program (and otherwise a professor of strategy and global e-business at the Robert H. Smith School of Business at the University of Maryland at College Park) has proposed what he claims is a new way to build big — the metacompany. It aggregates the key features of an incubator, a VC firm, and a diversified company. Like a corporation, a metacompany has a CEO and a corporate management team and maintains a significant, but less than 100 percent, ownership stake in a number of ventures, whether they remain closely held or later go public. But unlike VCs and incubators, metacompanies focus on a single area of business.

Mr. Gupta, who has spent 20 of his 50 years studying issues related to business diversification, says that he knows enough about the business to see that the new “meta” concept — as embodied so far by Raza Foundries and Comstellar Technologies — represents a significant twist on the old way of diversifying. The difference: he thinks the metacompany model could actually work.

The trick, Mr. Gupta maintains, is to focus the metacompany much more narrowly than the garden-variety incubator and narrower still than the typical Sand Hill Road VC firm. More focus makes for greater depth and specialization, which in turn should produce greater value, Mr. Gupta argues. We caught up with Mr. Gupta at Stanford to discuss the idea.

How does a metacompany differ from the obviously flawed incubator?

A typical incubator provides little more than space, commodity-type infrastructure services, perhaps some seed money, and a certain amount of generic startup guidance. A VC firm brings in serious money and, at the top end, provides a considerable amount of strategic guidance and access to managerial talent as well as collaboration opportunities. In contrast, a metacompany sees itself as an operating company that not only helps in the venture creation process, like the VC firm or an incubator, but can also help in the exploitation of collaborative synergies after the new ventures have become established operating companies. The goal of a metacompany is to reduce new venture risk by managing it rather than by diversifying broadly.

What, in your opinion, is wrong with incubators?

Let’s look at what happens in the case of most incubators. As a recent Harvard study has confirmed, the typical incubator provides little more than an identical array of commodity services. There is nothing wrong with being a provider of commodity services. They are needed, and somebody has to provide them. But why should providing commodity services entitle incubators to an average of 35 percent equity stake in their client ventures?

Given the dismal history of diversified firms, what makes you think a metacompany is going to be any different?

Typically, the promise of cost efficiencies that might emerge from diversification tends to be genuine. Too often, however, the attempt to realize these economies tends to be either inept or, over time, degenerates into ineptness. [One of the key elements of diversification is to centrally source common services and thus provide the benefits of scale economies.]

In reality, the incentive system in the typical diversified firm can almost never compete with the power of incentives embedded in a startup. Because of competition for scarce resources and promotions to higher levels, the diversified company becomes a hotbed of rivalrous competition [instead of] a collaborative network. So business units often prefer to coцperate with external third parties than with their peers inside the diversified firm. The centralized service units become bloated, slow-moving bureaucracies.

The metacompany sounds a lot like a diversified company, subject to the same kind of mess you’ve just described. What is the difference?

The metacompany is indeed an experiment in concentric diversification. In order for it to work, it must keep diversification narrowly focused on certain high-opportunity sectors of the economy, like optical networking equipment; ensure deep knowledge at the management level; stay lean and mean; cultivate but not mandate lateral collaboration; and be ready to disinvest when positive network effects have run out.

Comstellar Technologies is one example of a metacompany, focused on the communications industry. How is this company different from Harris Corporation, a vertically integrated communications company?

The trick is to look at Comstellar or Harris dynamically — as movies and scripts that evolve rather than as images frozen at a single point in time.

Depending on how it is managed, Comstellar can just as easily evolve into a Harris as it can into a Cisco Systems. Unlike Harris, Comstellar intends to own less than a 100 percent stake in its portfolio companies. This would not only permit the retention and cultivation of entrepreneurial incentives at the portfolio company level, but would also ensure that there would be mutual accountability.

Also, unlike Harris, where accountability ran only one way [from the units to the parent], in Comstellar’s case, the parent would also have to keep demonstrating to the other investors in its units that it is adding value. Comstellar’s agenda includes a heavy dose of new venture creation and not just the pursuit of operating synergies.

The odds are clearly more in favor of Comstellar. But, that’s all they are — odds. Comstellar still has to execute and deliver on its logic.

What about the entrepreneurs of new ventures? What should they consider before becoming part of a metacompany’s portfolio?

Remember the maxim: in life, the only thing more unfortunate than bad children is bad parents. Every new venture needs an ecosystem of essential resources. Joining a metacompany is like going to a full-service provider for one-stop shopping. This may appear convenient, but you must weigh the costs against the benefits. Do you need the specialized expertise that a metacompany might provide?

Is the metacompany that you are considering properly architected, and does it have the right kind of leadership? If the answer is “no” to any of these questions, then you might be better off with a good VC firm, or even a simple incubator.

Let’s assume that you are a metacompany, and a successful one at that. How do you make money? If you disinvest the portfolio too quickly, the network effect weakens. And, if you don’t, how do you show revenue and profits?

The right time to divest or disinvest from any company in the portfolio is when you can no longer add value to it, for example, Hewlett-Packard‘s decision to spin off its instruments business. Or if another parent can add greater value than you can, as in the case of Nokia‘s sale of its computer business to Fujitsu-ICL.

Some have claimed that it is possible to build a metacompany that might become as large as, say, Cisco or General Electric in five to ten years. Is that really possible?

While I would never say never, I would put the probability of a metacompany becoming as large as Cisco or GE in the next five to ten years as lower than that of, say, finding myself standing on top of Mt. Everest in the same time frame. [Editor’s note: Professor Gupta is no mountain climber.]

My advice to the leaders of a metacompany, to investors in them, and to the analysts who follow them would be to forget about size as having any relevance as a measure of success. The focus should be on return to investors.

When should a metacompany go public? Does one have to wait until a few of the portfolio companies get sold or go public? Or, can a metacompany go public much earlier on?

Like the question of when is the right time for two lovers to get married, the answer here depends very much on the readiness of the parties — the market as well as the company. The earlier in its life a metacompany goes public, the more investors have to rely on the promise of future success without actual concrete evidence.

Until April of this year, capital markets were indeed ready to place far bigger bets on promises than they ever had or are today. How times have changed! I think that, in today’s environment, a metacompany would have little choice but to build a track record of proven success before contemplating going public. In today’s type of market, I would tend to regard a desire to go public before having built a track record as a sign of desperation and, thus, a strong warning signal.

How do you value a publicly traded metacompany?

It is useful to think of a metacompany as analogous to a closed-end mutual fund where you as an investor have the option of investing in the portfolio companies directly. A metacompany that has become just a holding company of a nonchanging portfolio ought to sell at less than the sum of its parts.

By buying into the metacompany rather than the portfolio companies separately, all you get is reduced flexibility in reconfiguring your own investment portfolio. This is a cost that should be reflected in the form of a discount in the stock price of the metacompany.

On the other hand, a metacompany that is expected to keep morphing its portfolio is one that could potentially trade at a premium to the aggregated value of its components.