TechTalk: Don’t Pull That Plug! –

It was April 2001 and we thought we had seed funded a great new company. Savastra Systems Inc. was going to revolutionize the world of server architecture, making the task of adding servers far cheaper, easier and efficient.

“Silicon Solutions for Next-Generation Integrated Server Platforms,” crowed our PowerPoint. We would deliver, in typical tech jargon, “silicon and software building blocks to build scalable high-performance virtual server platforms for utility computing.” One patent had been filed, two were about to be filed and four more were identified for filing in the future.

We considered the concept a “can’t miss.” Rather than daisy-chaining servers as companies and data centers grew, we would eliminate redundancy and build a server system from the ground up with common memory, power supply and processors. And high-performance switch fabric would connect the whole thing, making it faster than traditional Ethernet backbone. All this would provide, in addition to substantial cost of goods savings, the ability to offer better management software and scale systems linearly as demand grew. That would afford, by our projections, a 10x advantage over existing systems.

Despite the nuclear winter that held its chill over Silicon Valley, we were willing to take a risk on the technology by seed funding the company. If the market were as prescient as we hoped it might be, data centers would soon be clamoring for Savastra software and hardware.

Unfortunately, we were 18 to 24 months ahead of the market. Customers didn’t want leapfrog technology; they wanted incremental change as they continued to cut costs and squeeze more productivity out of existing technology assets. They would buy blade servers for incremental growth. New server architecture would have to wait.

Having invested close to $1 million in a company with fewer than five employees; no outside investors, board members or creditors; and technology that could sit on a shelf for a year without getting moldy, we decided to mothball the company. It’s an option few VCs consider but, under the right circumstances, they should.

A Viable Option

Mothballing isn’t a common term within the venture capital lexicon. It is often a reactive rather than proactive decision. I would argue it should be just the opposite. VCs should not simply view the shuttering of companies as their only option when deciding whether to pull the plug or invest more money. Nor should they consider mothballing an afterthought before the final board vote in making these decisions.

Rather, carefully chosen portfolio companies can be proactively steered toward being mothballed before racking up liabilities or burning through excessive amounts of limited partner capital. It should be a board level decision as important as whether to plow more money into keeping a company alive or allowing it to simply run out of cash.

“The board has to face the question of, Is there some time in the future when there’s a decent chance to capture greater value for shareholders than could be captured now without incurring any excessive costs in doing so?'” says Paul Graffagnino, a partner with GCA Law Partners of Palo Alto, Calif. “If the answer is yes, then you should mothball a company. But it’s a big decision for VCs-as much a psychological decision as it is a business decision.”

Graffagnino is right. It is as much a psychological decision as it is a business one. But in the past, VCs have traditionally allowed the former to outweigh the latter without doing the math. The hassle of having a company sit on the shelf-of employing one or two employees to monitor the market, keep IP updated, pay any licensing fees and file any necessary corporate paperwork-seemed far more burdensome than simply shuttering a company and selling its assets.

Indeed, Monday morning partner meetings hardly need to keep track of mothballed companies on top of the living dead already littering most portfolios. As a result, the psychological argument against mothballing is a strong one. What it shouldn’t be, however, is a defining one.

Is It Worth It?

“Mothballing is essentially an option on the IP,” says David Kirsch, an associate professor of management at the University of Maryland Business School. “It’s a way of extending the life value of a company which has declining net present value.” If venture capitalists consider the cost of the option as essentially being the value of money already invested, the business and financial argument gains traction-assuming that several important conditions are met.

First, the IP must be capable of having a certain shelf life. Biotech and software are easier to store simply because the markets for diseases and software solutions aren’t likely to go away, whereas hardware or communications markets are often more dynamic, changing rapidly with incumbents and new entrants offering a higher velocity of solutions.

Second, mothballed companies must be at least financially capable of sitting on the shelf unencumbered by creditor liens. In other words, assets must exceed liabilities at the time of mothballing. No creditor is going to allow a company to be stored in order to capitalize on some future value of its IP when the company owes it money. In these instances, says Graffagnino, “sell the assets as fast as possible to the nearest buyer and shut it down.”

Third, mothballing companies requires fairly unanimous board approval and minimal pushback from a VC’s own partners. This can’t be overstated. For VCs, there’s a certain amount of machismo in not wanting a living reminder of a failed investment (at least in the short term.) For entrepreneurs, who are experts at being unrelentingly optimistic and particularly bad at being brutally realistic, mothballing may be even worse than shutting down a company. Though it may keep their dream alive, it may be a dream that founders may no longer be involved with in the future when the band is put back together.

Finally, a company should be mothballed only if the IP can be preserved for future business and technology purposes while any existing customer base can be preserved without destroying goodwill. In other words, if a modest amount of staff can be employed to service existing revenue while maintaining the relevance of the intellectual property during the time it is sitting on the shelf, mothballing makes sense.

When and if a company is revived down the road, the IP cannot be so irreconcilable with the engineering talent that created it that the value of what was shelved becomes meaningless at the very point it should hit the market. In practical terms, mothballing Savastra will only bear fruit if its patents, architectural plans and software code can be easily revived when and if we choose to breathe new life into the company.

Buy-In Required

Once the decision is made to mothball a company, there are a couple of steps that VCs need to follow.

First, board members and key employees need to be given some incentive to keep them at least mindful of the future success of the company. In the case of Savastra, Krishna Viswanadham, a Charter entrepreneur-in-residence who has helped guide the company, is expected to keep approximately one-eighth of his mind focused on the server platform market during the time when the company is mothballed. When the market becomes more available to Savastra’s technology, Krishna and others will play a strong role in restarting the company, while still having equity tied to its potential success.

Second, companies must be mothballed in such a way such that VCs can easily hit “Ctrl-Alt-Delete” to restart the system. In this sense, mothballed companies might require a modest amount of funding to bridge them through the next one to two years. We would estimate roughly $10,000 per month-no more than $50,000 per month-in seed money to time capsule a company, keep its IP fresh and potentially maintain one business person to keep any relevant sales cycle alive.

From a legal perspective, the costs of mothballing are minimal. Although psychologically it might be cleaner for VCs to not have anything leftover from the last several years hanging in the closet, in practical terms the legal fees for keeping a company incorporated are roughly only a few hundred dollars. Given the choice between eating several million dollars in funding in a shutdown or paying a minimal amount in the hopes of resurrecting good technology when the market swings more to our favor, we are going to actively choose the latter.

For those venture capital skeptics who would rather keep their decisions simple, there’s no question that simply shutting down a company or opting to fund it one more time is an easier decision. But if VCs are willing to simply calculate the net present value of the IP now vs. the future, without necessarily having to raise another full venture round to keep the IP alive, mothballing indeed represents a third viable option we should all start considering.

The most important question, however, is the one most often forgotten. Does mothballing a company really work? “I’ve seen them work,” says Graffagnino. “They’re definitely in the minority, and no one wants to speak about mothballed companies too loudly, but I’ve clearly seen a few of these companies come back.”

With the amount of capital we’ve already spent on technology that has yet to see the light of day, it seems only right that before throwing good money after bad, we proactively consider mothballing companies before shutting them down. In fact, maybe we should even mothball a few before they burn through the remaining cash they still have in the bank!

Ravi Chiruvolu, a general partner of Charter Venture Capital, is a regular technology columnist for VCJ. If you’d like to send him feedback or ideas, email him at Chiruvolu specializes in enterprise software, software infrastructure, e-business and wireless technologies. He sits on the boards of Ellie Mae, ManageStar, Quantum3D, Talaris, Verano Winery Exchange and Xavient Technologies.