Innovation is more than invention

We thought we were home run hitters, but then we learned that we were born on third base. — Attributed to an AT&T alumnus

This quote by Stanford economist Tim Bresnhan (at Tuesday‘s State of the Net West conference) nicely captures the key problem facing innovative engineers working for any company smaller than the old Ma Bell.

Bell Telephone Laboratories had a better research staff (and working conditions) than all but a handful of universities. The monopoly profits of The Phone Company fueled some of the greatest inventions and scientific discoveries of the latter 20th century — things like lasers, satellites, information theory, etc. etc.

However, as this quote suggests, there are two reasons why such inventive output doesn’t provide a good measure of success:

• It’s easy to bring technology to market (e.g. electronic switching, microwave long distance) if your captive customers comprise 90+% of the largest economy in the world.
• When you can’t bring your expensive new technology to market, the guaranteed rate of return means there is no penalty for waste or inefficiency.

Invention is never enough. Except at Ma Bell, big innovations require (as Hank Chesbrough demonstrated in his Xerox studies) a business model and someone entrepreneurial enough to create a market if it doesn’t exist already.

Killing a business

Friday’s paper brought news of the planned liquidation of the Smith & Hawken, the premium garden store founded in 1979 in Marin County. Although I’ve never visited one of their stores, it was nonetheless sad news, because the company provided important inspiration for my early days as a tech entrepreneur.

When we founded our company in July 1987, my partner Neil Rhodes and I were intentionally (and ultimately unsuccessfully) emulating Hewlett & Packard. Like Bill & Dave, we started in a garage, and we deliberately emulated certain elements of the HP culture (which also became our biggest customer).

However, it was Smith & Hawken that provided explicit validation for some of our ideas about what a business was and could be. A year after we launched, in 1988 we read Paul Hawken’s Growing a Business; this was decades before the two comparable HP histories, The HP Way and Biil & Dave.

In his book, Hawken talked about issues that resonated closely with us, things like how to treat employees and the meaning of business. Although it’s been 20 years since I opened the book, I recall his view as being that creating and growing a business was more than just about money, and that founders could and should shape the firm to reflect what they believed important.

Although his industry was very different, this was a view quite consonant with tech startups of the 1980s, back when people tried to build a business for the long haul and before IBM and HP layoffs changed forever the idea that somehow tech companies were different.

Perhaps Smith and Hawken weren’t different, either. Hawken used the success of the book and the company to branch out into other things, leaving the company in 1992, which was then sold in 1993 for $15 million to the parent of NordicTrack. Some of the experiments by the founders (losses on an unsuccessful clothing line) were seen as contributing to the company’s need to be acquired

Since 1993, Smith & Hawken has had three corporate parents. The last owner was Scotts Miracle-Gro, which bought the company in 2004 and has been unable to find a buyer for the firm. Some of the turnaround moves sounded plausible, but nothing worked.

The two founders showed none of the remorse one might expect of a parent outliving its child:

"Scotts couldn't have been a worse corporate owner," said Hawken, who lives in Mill Valley. "Smith & Hawken had become just a ghost of itself."

[David] Smith, who lives in Mendocino County and owns Mulligan Books in Ukiah, said he had gone so far as to ask friends not to shop there.

"When Scotts bought it and Smith & Hawken was owned by the largest pesticide seller in the U.S., I suggested people boycott it," he said. "It had completely lost its roots."
…
Hawken, chairman and CEO of engineering company Pax Group, used the occasion of the closing to host a party Wednesday night. "I couldn't be happier to see my name come down," he said.

What’s the moral of the story? Perhaps that nothing lives forever. Certainly that (as my mentor Charlie Jackson advised me 15+ years ago) once you sell a company, you have to let go and accept what happens. Also, people change — even visionary founders — and their passions will also change.

Most of all, I think it says that even with a unique business concept and vision, firms need to have the management depth and the dispersed leadership and the culture to continue on after their founder is gone. This is certainly a crucial issue for Apple to deal with, more than a decade after Steve Jobs undid the terrible damage done by his three CEO predecessors, and nearly five years after his initial cancer surgery.

Is entrepreneurship down, and why?

Scott Shane of Case Western argues that entrepreneurial activity is declining, due to competition from larger rivals with superior scale.

Meanwhile, Jonathan Adler (also of Case Western) has a contrary argument:

It seems to me that another likely contributor is the increased regulatory burden. It is well documented that regulation can increase industry concentration. Smaller firms typically bear significantly greater regulatory costs per employee than larger firms (see, e.g., this study), and regulatory costs can also increase start-up costs and serve as a barrier to entry.

What I find interesting is that entrepreneurial opportunities and activities are not equally distributed across the country: more firm formation happens per capita in Silicon Valley than in (say) Northeastern cities. And California is larger than any other state, the total numbers will be even higher than the ratios.

Today, California is going through a series of changes that are making the business climate less attractive, especially for startups:

• a shortage of venture capital
• increasing taxes
• cutbacks in education and other infrastructure spending
• a political climate that is encouraging increased regulation

California entrepreneurs may continue despite such burdens, or they may move somewhere else, or they may not start a company after all. But a decline in entrepreneurship in California would certainly reduce the total number of US startups more than in a smaller or less entrepreneurial state.

Panel on bootstrapping tech companies

Not surprisingly given the membership pool, the MIT Club of Northern California has four activities related to technical entrepreneurship. This includes separate speaker series on semiconductors and clean technology, as well the Venture Mentoring Series for alumni (more at a later date).

One that I didn’t know about is the C3 group (Convergence, Community and Commerce), which put on a panel last month called “The Art and Science of Bootstrapping.” The four bootstrapping panelists were

• Paul Kocher, Founder, President and Chief Scientist of Cryptography Research
• Michelle Munson, CEO and Co-founder of Aspera Software
• Beatrice Tarka, CEO and Co-founder of Mobissimo
• Sridhar Vembu, CEO and Founder of AdventNet

Although I didn’t make it, there is a long report in the Spring 2009 edition of the MITCNC newsletter.

Not surprisingly, the interest in bootstrapping is motivated by the increasing risk aversion of VCs, making it difficult to get seed funding. Entrepreneurs are having to bootstrap off of personal (or 3F) money to grow the business to a point where it’s interesting to VCs.

There are two paragraphs that caught my attention:

All panelists agreed that a company cannot always remain bootstrapped and there are circumstances when the founder should be open to accepting outside financing. Paul Kochar further elaborated that if a company is in a market that is poised to takeoff and there are no barriers to entry for other competitors to come in, moving fast is key to capturing market share. In such circumstances when a company can grow very quickly, it is imperative to look for the type of financing that can facilitate such growth, for example, to develop customer support and sales organizations or to invest in user experience. On the other hand, if a company is in a niche market with no competitors, then one can afford to bootstrap and slow the growth to maintain financial control.

I am guessing from the report that all of the entrepreneurs had high growth aspirations — and hence their concern about growing quickly before the window of opportunity closed. Here is the other paragraph:

Regarding how long a company can be bootstrapped, Michelle said that it is easy to bootstrap in the first two years of a company’s life, but beyond that it was less clear because the window of opportunity remains open only for so long. Sridhar said that if one does not have a big idea and no big discontinuity, it is better to be bootstrapped. Only 1% of new businesses are venture financed. Paul pointed out that some arenas, such as clean tech or semiconductors, are not generally good candidates for bootstrapping.

Actually, 1% is much too generous a number. By my own calculation, funding peaked in 1999 (during the dot com bubble) at about 0.15%, but in the early 2000s averaged about 0.037% — one company out of 2,700.

Obviously some of the other 2,699 companies are those (like my own) that are high-tech startups or have a strong technology component. Many of these are bootstrapped forever, but this normally consigns the venture to slow growth niches.

This is how I would have summarized the event if I’d attended:

Bootstrapping is a way of getting launched and proving viability to outsiders. However, if you are pursuing a rapid growth opportunity, then you will soon need sizable outside funding — either VC or corporate VC. If your competitors have it and you don’t, you will be left behind.

The greatest source of tech entrepreneurs

Journalists love anniversaries, particularly ones with big round numbers. Politicians (and others seeking publicity) love them two. They’re what we call a “news peg.”

On Sunday, the San Jose Mercury published a long article by Scott Duke Harris celebrating the 100h anniversary of the founding of Federal Telegraph and Telephone in Menlo Park. The article mentioned two other Stanford-related startups founded in the first half of the 20th century — HP [founded in 1937] and Varian Associates [1948] — as well as the familiar list of IT firms from the 1980s and 1990s.

Harris used this history to make a point about Stanford’s role in promoting high-tech entrepreneurial culture:

Stanford University's 100-year tradition of entrepreneurialism, which has spawned such tech giants as Hewlett-Packard, Cisco Systems and Google, has been recognized as a catalyst to Silicon Valley's emergence as the globe's pre-eminent tech hub.

Alas, to quote Will Rogers, “It isn't what we don't know that gives us trouble, it's what we know that ain't so.”

OK, I’m a little biased here. I’m an MIT grad, active in the local MIT club, and onetime entrepreneur alum. I have been researching MIT-trained entrepreneurs as part of a book tentatively named From MIT to Qualcomm. There are also key Stanford rivalries, with strong personal and family ties to the UC system, and today teaching tech entrepreneurship in the shadow of the world’s second richest university. (Full disclosure: I was accepted by Stanford the only time I applied, as a high school senior).

And by no means do I want to minimize the role that Stanford has played in sparcing alumni (and faculty) entrepreneurship in Silicon Valley, starting with Cypress Semiconductor, Electronic Arts and Sun Microsystems in 1982, and extending through Cisco, Yahoo and Google (among many others). After benign neglect by the business school, Stanford’s engineering school has played an incomparable role in promoting technology entrepreneurship among students at Stanford and elsewhere, with efforts like the Stanford Technology Ventures Program and its free iTunes U podcasts.

However, after talking to Silicon Valley historian (and Palo Alto native) Stephen B. Adams, I believe claims of Stanford’s role in the early 20th century are greatly exaggerated. As Steve wrote me in an email responding to the Merc article:

The early start-ups (pre HP) did not take because the Valley lacked critical mass of high tech talent. Therefore, local firms such as Federal and Farnsworth moved to areas (New York/New Jersey and Philadelphia respectively) that already had clusters going. The 1939 Census of Manufacturing showed the same thing that Fred Terman later said: by then the Valley had fewer than 100 scientists and engineers in industry. Not exactly critical mass!

In fact, there wasn’t much to HP until it wartime orders swelled its ranks to 200 employees, and it laid off more than half of those after the war.

At best, Stanford’s role as an entrepreneurial incubator began when Fred Terman was appointed engineering dean in 1946, or when Stanford decided in 1951 to allocate some of its land to form the pioneering Stanford Industrial Park. Even in the 1970s, Stanford’s role in creating startups was not obvious. Shockley (and then Fairchild and Intel) put the “silicon” in Silicon Valley without direct ties to Stanford.

Adams points out that Stanford was aggressively ahead of Berkeley for a very simple reason: without Berkeley’s ongoing support from Sacramento, Stanford badly needed the money. Necessity is the mother of invention.

Which brings me to the other point. As Anna-Lee Saxenian documented in her 1994 book, Regional Advantage, Silicon Valley surpassed Route 128 in the 1980s and never looked back. Saxenian says it’s because of the valley’s open culture, but others argue that it’s because Boston bet on minicomputers while the Bay Area bet on PCs.

However, in the period 1920-1970, there was no question which university was inspiring and fueling technological entrepreneurship: it was MIT.

Losing its land grant status and rejecting a proposed merger with Harvard, in the early 1920s MIT was scrambling to raise resources both for the Institute and for its faculty. (Remember, big Federal R&D spending started with WW II).

During the interwar era, MIT invented its industrial cooperative program (allowing students like Andy Viterbi to work in real jobs to pay for their schooling). It also invented the now-standard consulting rubric used by all American research universities, the “one day a week rule.” This is well recounted by Henry Etzkowitz and his fascinating book, MIT and the Rise of Entrepreneurial Science.

Through the end of the 1960s, MIT was also the world’s leading university in creating the field of electrical engineering and establishing computer science as an academic discipline. MIT has its own list of alumni- and faculty- (co)founded companies, with an impressive run of electronics-related startups from 1922-1985 in both Massachusetts and California that included Raytheon (co-founded in 1922 by Vannevar Bush), EG&G, BBN, TI, DEC, Bose, Lotus, PictureTel, 3Com, and Qualcomm. Among those launching its many Bay Area startups were the two men who put the Silicon in Silicon Valley: William Shockley and Robert Noyce.

By the way, where did Terman (son of a Stanford psychologist) learn his trade as a radio engineer? As one biographical article recounts:

Stanford’s own Electrical Engineering Department chairman told Terman that the biggest and best EE department in the country was at MIT. So in 1922 Terman joined a generation of promising EE graduate students on a pilgrimage to Cambridge.

At MIT, Fred undertook graduate study under Vannevar Bush. Fred earned his Doctorate Degree in electrical engineering in 1924, and having a fascination with all of the exciting events at MIT, Fred was to accept a teaching position there.

Due to health reasons, Terman returned to Stanford, but he spent 1941-1945 running the Harvard Radio Research Laboratory, an auxiliary to the much larger MIT Rad Lab.

Today, both MIT and Stanford have exceptionally qualified faculty, undergraduate and graduate students in engineering and the sciences. MIT’s technical role remains strong, but certainly (as the Merc argues) Stanford has taken the lead in fostering tech startups. Stanford has a much better local environment for entrepreneurs, but it’s an open question whether that’s due to West Coast vs. East Coast cultural differences, the availability of VCs, or the wealth of successful entrepreneurs (and entrepreneurial wealth).

If MIT has been eclipsed by Stanford in firm creation, there’s no guarantee that the latter will remain pre-eminent indefinitely. Stanford’s main rival will not be Cal (or even MIT), but instead Tsinghua or the various campuses of IIT.

I don’t think it will happen right away: these other schools may be able to imitate Stanford’s talent, but it will be longer before than can replicate the ecosystem that lies in its back yard. So even if the Merc has the history garbled, the contemporary story of Stanford’s entrepreneurial success is true (at least for the time being).

End of the IPO anomaly?

In 2008, there were only six IPOs nationwide, compared to 365 twenty years earlier. The end of IPOs means that founders no longer run their companies, but instead get acquired by big firms and then quit to do something else.

Sunday’s Merc, presented its annual compilation of the Silicon Valley 150. One noticeable result was a reduction in the number of public companies and also of IPOs.

Columnist Chris O’Brien remarked on the trend:

From 2001 to 2008, there have been 90 IPOs in the valley, an average of 11 annually — and the last one was more than a year ago. Compare that with 331 IPOs in the years from 1990 to 1998, an average of 41 annually. The two years in between were so insane — producing 163 IPOs — that it's no use considering them for sake of comparisons.

We’ve long known that Silicon Valley has a higher rate of IPOs than in other countries, but the evidence also suggests it has a higher rate than the rest of the US.

In my own research on communications startups in San Diego, I’ve noticed a much lower rate of IPOs. By my most recent tally, there are eight public companies in the telecom cluster (a few were either acquired or died after their IPO).

So, I’m sorry to say, the data is starting to confirm my conjecture. The 1980s and 1990s offered an unusual window of opportunity for IPOs — both in terms of the availability of financing and the ability to create a new stand-alone company.

Acquisitions do have a few advantages: they are quicker, available to a wider range of firms, and less dependent on the cyclical capital markets. My mentor Charlie Jackson planned for an IPO but sold his company to Aldus in 1990 when the IPO market closed.

I wonder when (or if) the business (or engineering) school courses in entrepreneurship will notice the change, and adjust their curriculum accordingly.

Illegitimate startups

In grading business plans this week, I was struck by a common blind spot: my undergraduates were overly optimistic about their chances of gaining sales (or distribution) from day one.

They didn’t realize that they would be handicapped by the inherent lack of legitimacy that a brand-new firm has. This is something I lived as a software entrepreneur, but also a well-known problem to strategy researchers.

Just as innovation and entrepreneurship scholars often trace back their core theory to Schumpeter’s “gales of creative destruction,” those worried about the disadvantages faced by young firms go back to sociologist Arthur Stinchcombe and his 1965 book chapter which coined the phrase “liability of newness”.

Subsequent research (such a 1986 ASQ article by Singh et al) has shown that the liability stems in large part to the lack of external legitimacy held by the new organizations. (The test was with nonprofits, but the principle is the same).

I had trouble finding something suitable for undergraduates to read; perhaps this is a publishing opportunity. However, I but did find a closely related 2008 article in Entrepreneur entitled “Credibility is King.”

In class, I tried to tease out the difference between legitimacy and credibility, but given the impromptu nature of the discussion I have to admit I was mostly winging it. In my mind, legitimacy is whether or not you’re in the consideration set. Credibility (as in the political context) is whether or not customers believe what you say.

We brainstormed the reasons or conditions for a lack of legitimacy, and I noticed that they all boiled down to two issues. One is the lack of information (e.g. about the unmeasurable quality of your good). The other is for cases of high risk: making a bad decision on a bicycle helmet is different than doing so for buying an ice cream cone. Again, this is something I hope to elaborate in a future article.

In terms of solutions, most of our options boiled down to two. One is to borrow legitimacy, such as from suppliers (“Intel inside”), dealers, or testimonials by buyers or celebrities. The other is to reduce the risk faced by buyers, such as by providing a free trial.

Again, this is something I’d like to write up sometime, but at least I have a starting point to sensitize future students to this challenge they will face as entrepreneurs.