Tuesday, October 31, 2006

Eating Less: The Elixir of Life

Juan Ponce de Leon traveled to the new world in search of the fountain of youth - today, the NY Times Science Section reveals the secret to extended life: caloric restriction.

In a study involving rhesus monkeys, one group of monkeys received ~50% fewer calories than another group - 445 vs 885 per day. The results are startling and "cast doubt on the long-held scientific and cultural beliefs regarding the inevitability of the body's decline."

The well-fed animals suffer from obvious signs of age, relatively more frequent incidents of diabetes and cancer, and a higher mortality rate. For humans, caloric restriction would involve reducing, on average, daily consumption of calories from 3,000 to 2,000.

Scientists are working on mechanisms other than all of us becoming ascetics to provide similar results. The story discusses an interesting experiment with earthworms whereby a mutation of the gene daf-2 led to a 6x increase in life span. The gene appears to "trick" cells into thinking that nutrients (insulin) are not available and in turn prolongs the live of each worm cell.

Increasing insulin and calorie consumption appears to accelerate cell death and reduce life span - I suppose the next time we drive by In-N-Out Burger, we will need to decide if that double-double is worth it after all!:)

Building an Open Source Business

Dave Rosenberg, CEO of Mulesource, recently published an informative article on Sandhill.com regarding his experiences in starting an open source company - from looking for funding, to legal and IP issues, to community development, and pricing, Dave's article is good food for thought.

He describes MuleSource as a second-generation open source company - that is an apt description as RHAT, JBoss, MySQL and others have paved the way with respect to business models, enterprise acceptance of open source, sales strategies, and experienced open source employees are accelerating the rise of recently funded companies.

Thursday, October 26, 2006

Trial and Error

Start-ups are, if anything, exercises in iteration. Hypotheses are formed, tested, adapted, retested, and a given company's center of gravity, culture, and path are forged in the fire of trial and error.

As an early stage venture investor, it is interesting to think about the minimum number of iterations required before a company is sufficiently baked that one's capital will be used to refine, rather than to reinvent, a given business.

Refinement and course correction are fundamental to start-ups, reinvention, however, is more related to inefficient capital consumption than it is to value creation.

I often state that genius is a function of context. The best ideas and the most valuable innovations germinate from real world observations of customer need.

Said another way, projections are dangerous, while rejections are instructive.

Start-ups, without any a priori knowledge of the customer need/problem, often project need and develop solutions in isolation and without the bearing of market reality - similar to Plato's Allegory of the Cave. It is always best to ensure that inputs are a function of reality (light of the sun) rather than projection (shadows on the wall).

My personal view is that the minimum number of iterations may be hard to quantify, however, I have observed that the best time to meet a company is ~T+6months. 6 months of commitment, iteration, and market reality seems to be the optimal balance between opportunity and trial and error.

The maturity of the plan, product spec, pitch, value prop, team's commitment, etc is exponential across time - the job of the early stage VC is to work out how much time and how many iterations and if a given company presents the optimal mix of both factors.

Wednesday, October 25, 2006

Decision Making and Uncertainty

Investing and management are in many ways exercises in decision making.

Where to invest? Where to focus? What are the risks? What is upside, the downside, and what to do and, just as importantly, what not to do?

Thoughtful investors and managers work hard to analyze the pros and cons of a given decision, however, no matter how diligent the analysis the final decision will be a judgment call that involves a large degree of uncertainty.

If imperfect knowledge is a given and yet decisions must be make, what to do?

Bob Rubin is well known as a legendary arbitrage trader, co-CEO of Goldman Sachs, and Secretary of the Treasury. His wonderful autobiography In an Uncertain World provides important insight and lessons into how to best make decisions under uncertainty. The governing principle of his book is that nothing is provably certain.

He writes,

"In arbitrage - as in policy making - you also have to be able to pull the trigger, even when your information is imperfect and your questions cannot all be answered. You have to make a decision: Should I make this investment or not? You begin with probing questions and end having to accept that some them will be imperfectly answered - or not answered at all. And you have to have the stomach for risk."

Rubin does not believe that given the certainty of uncertainty that decision making should be from the gut or based on intuition. Rather he argues that decision making is fundamentally about calculated risks. He stresses a focus on defining a set of possible outcomes, quantifying the probabilities and payoffs associated with each outcome, and then quantifying the expected value payoffs for each outcome identified.

As an arb trader and policy maker, Rubin focused on identifying the potential upside and downside risk, the net of which is the expected value. For example, at GS he looked to take risks that would generate a return of no less than 20% on the firm's capital.

Similarly to Naseem Taleb's book Fooled by Randomness, Rubin stresses that good decisions that appropriately weighed the pros and cons can and will have bad outcomes. Also, it is clear that bad decisions can have good outcomes and that we can be fooled into believing that given the successful outcome that causality is clear - ie that we made a great decision.

Taleb's book centers on the "hidden role of chance in life and in the markets." As an investor, it is particularly apropos as one tries to identify systematic methods of creating value via investing. Investors and investees like to believe that the investment world is deterministic with clearly understood cause and effect. Understanding causal drivers of value helps to create repeatable models for investment that scale both across time and individuals in the firm. Nassim challenges us to be very careful in overascribing reason and logic to an outcome. Too often, investment outcomes are the result of randomness rather than science, ie being lucky rather than good.

He illustrates his point by comparing the world of investors to a large sample of coin-flippers. If you start with a large enough sample, someone will manage to flip heads for many months in a row. While others drop out, the surviving "flipper" will take on magical qualities and others will study his background, methods, and secrets for achieving such success. Then, of course, he will flip tails one day and "blow up."

Nassim's book instills a deep sense of humility in the reader and a recognition that time and chance are often at the root of success. I strongly recommend reading his book.

Both Rubin and Nassim provide valuable insights into how best to operate under uncertainty. There is much to learn from both books.

Successful people and firms focus on building frameworks that allow for systematic decision making and judge people ultimately on the quality of the analysis and the assumptions used to derive at a decision rather than simply on the outcome itself. We all live with uncertainty - how we make decisions day after day given that fact will separate the winners, the losers, and those simply who got lucky!

Thursday, October 12, 2006

Infopia and Salesforce.com

Congratulations to Infopia, a Hummer Winblad company, for winning the inaugural Salesforce.com's Appy Award Breakthrough App of the Year!

Infopia provides multi-channel online selling solutions. Infopia's Marketplace Manager solution empowers sellers to reach all possible sources of liquidity and transactions via a single merchandising, inventory, shipping, and complete order to cash system. The integration with Salesforce extends CRM records to include a customer's transactions not only on the retail site, but importantly across all marketplaces, such as eBay. This single view of the customer is a critical extension to CRM and reflects the growing importance of marketplace transactions for retailers.

Congratulations to the team on another big win!

Monday, October 09, 2006

Poison the Well

Throughout history, retreating armies have poisoned wells, burned crops, and otherwise left the victors with barren lands. Which brings me to Sevin Rosen.

Not simply content to quietly shut down after nine funds and several decades of very succesful and significant investments, the firm announced the industry broken and incapable of creating value for its limited partners. In articles in the NYTimes and interviews on CNBC, Steve Dow is making his case for an industry in decline doomed to poor investment returns.

I must say that I find it remarkable that a set of investors whose professional existence is predicated on funding disruption and innovation are publicly advocating that an end-state in the venture capital industry has been reached; that no viable models exist, no innovations exist, no disruptions exist that will allow people to rewrite the venture rule book and add value.

There are no end-states, but rather constantly changing constraints that demand adaptation to ensure survival in the brave new world.

In June 2005, Howard Anderson, a co-founder of Battery Ventures, wrote an article titled Goodbye to Venture Capital in which he lamented how too much money and companies combined with stagnant tech spending had slayed the golden goose. It has, frankly, become cliche to discuss structural challenges and to see nothing but poor risk adjusted returns on the horizon.

Strategies, however, reflect the context facing any given business. The context today is quite clear - and commented on by me here - surplus capital, relatively poor IPO market, and too many venture firms/people.

In the face of stark constraints, the challenge is to define an intellectually credible strategy for creating value that incorporates today's realities rather than ignores them. At Hummer Winblad, we focus on capital efficient software companies priced to reflect the reality that the median IT exit is sub $100m. In the ROIC ratio, the denominator is fixed by the investor and entrepreneur. If the market is fixing the numerator (iemedian exit<$100m), then the maximum amount of money invested must take the exit as a given and be as low as possible.

Business models, use of proceeds, total capital required to reach profitability, and the weighted-average post-money valuation must all reflect the new reality. Steve Dow argues that, "Maybe we have to look only at deals that are going to take a limited amont of capital." I would argue that is logically consistent with simple math: exit/investment = return on invested capital.

Adam Smith famously stated that specialization is a function of market size. Venture capital is a massive market with significant specialization. There is no, accordingly, single model of venture investing and therefore no way to comment on the model failing. Certain firms and certain strategies that do not reflect the empirical constraints of the industry and adapt and innovate will certainly underperform.

My read is that firms will need to go earlier to avoid the ugly exit realities, or they will go abroad to leverage growth markets. Mid to late stage US investing will be a real challenge. The US will see smaller funds that are economically viable in the current market.

Life is disruptive and all business people - operators and investors - need to constantly question their strategies and demand an intellectually credible answer to how to best compete given the exogenous variables at work in any given industry.

VC is no different and it is too large and specialized an industry to extrapolate industry malaise from the hardships of a few funds. While I have great respect for Sevin Rosen's track record, I am not quite sure why going on CNBC to discuss the "broken" industry is a good idea.

Thursday, October 05, 2006


Jason Maynard, the lead software analyst for CSFB, published today a research brief on the significance and potential of Mulesource. His key comments follow: a powerful endorsement of the company and opportunity.

The Mule is Kicking

· Earlier this week MuleSource, a startup enterprise service bus provider, made its public launch as "the open source choice for integration and SOA." MuleSource raised over $4 million in venture capital funding from Hummer Winblad Venture Partners and Morgenthaler Ventures. MuleSource is headed by CEO Dave Rosenberg and CTO Ross Mason.

· We think this is an interesting company to watch since the ESB is the foundational layer for SOA and MuleSource has perhaps the most mature and widely adopted open source ESB. To date, Mule has more than 200k downloads, a community of over 500 developers, and is in production with over 100 organizations including several in the Fortune 50. In addition, the overall market being addressed by MuleSource is relatively big ($8.5B in software license and $132B in services according to Gartner), which should provide good opportunity for the company to grow.

· The early deployments of Mule have been for SOA and integration at the edge of the network. A few of the more advanced customers have progressed quickly and are now running the product as an enterprise backbone including a few transaction-heavy processing environments in the financial services market. The company hasn’t publicly disclosed any of its customers but we have heard good feedback about the product from many users. Mule’s business model is to offer subscription support contracts in a similar manner to other open source providers.

· Mule is designed to be open and agnostic with nearly all leading application servers, business process management tools, registry offerings, and standard security frameworks. One argument for open source SOA is that it can help solve some of the mundane and one-off architectural challenges for integrating niche applications and systems. The company has been pleased with the value-add from the community around developing new integration adapters. So far nearly 30% of the community is active in offering bug fixes and domain expertise around integration modules.

· One interesting trend to watch is the uptake of the SMuT stack (Spring, Mule, Tomcat) as an alternative to traditional application servers. Their positioning around the edge could siphon some business from the application server market and pop up as a potential alternative to other SOA offerings from JBoss (Red Hat), BEA’s Aqualogic, and TIBCO’s Matrix offerings. Mule is not positioned directly against incumbent messaging providers like TIBCO or IBM since it will take some time and proof points to validate their capabilities. Mule is still very in its lifecycle but should grow quickly given strong usage and product adoption. We don’t think this will disrupt business for the established players in the next year or so but it is an interesting seed that could alter the space in the longer term.

Monday, October 02, 2006


The inevitable march of open source software up the stack continues with Hummer Winblad and Morgenthaler's investment in MuleSource, reported here by Cnet's Stephen Shankland.

MuleSource develops and supports Mule, the leading open source Enterprise Service Bus (ESB) and Integration Platform. At funding, the Series A company is blessed with an active developer community (200,000 downloads), Fortune 50 enterprise deployments, and the enviable value proposition of superior, standards-based products available at a fraction of the cost of commercial source alternatives.

Mule is a Java-based platform that enables enterprise developers to perform a wide variety of integration tasks, from bringing new applications into production, to modernizing legacy applications and platforms, to enabling SOA (Service-Oriented Architecture). Mule's programming model enables faster integration results than proprietary solutions, at a fraction of the cost. And unlike proprietary approaches that frustrate users with complex, closed frameworks, specialized skill sets and architectural lock-in, Mule’s modular design enables enterprise developers to take integration into their own hands.

Enterprise IT organizations must grow IT capacity in a non-linear relationship to revenue and transaction volume growth. The move to SOA and composite applications requires that IT organizations make systems and applications available to an ever widening array of consuming systems and applications. Integration, traditionally a source of high-costs, must provide operating leverage -ie support the capacity of increased integration without a linear increase in the cost of integrating the nth system.

RDHT, MySQL, and others offer a mechanism of adding capacity at greatly reduced costs - keeping IT budgets flat to 1-2% up while adding 30-50% incremental capacity.

MuleSource is fundamentally aligned with IT's mandate to be a source of operating leverage for the business and it will be a great company to watch.

Also, the CEO Dave Rosenberg is an active blogger and co-writes InfoWorld's open source blog. Check it out.