db4o Developer Community

I had blogged and spoken about the inherent incompatibility between Open Source businesses and VC funding previously and we have sticked to this assumption all along. Successful people as different as Linus Thorvalds and Mark Leslie, founding CEO of Veritas and our chairman, subscribe to the notion, that open source is a marathon, not a sprint.

Under this assumption, db4objects has now been handed out a seven-figure personal check from SUN co-founder Vinod Khosla, one of the most prominent guys in Silicon Valley and repeatedly ranked by Forbes' Midas list as the most successful investor worldwide. Vinod is not only a visionary in technology, energy, education, and the emergence of India and China, but also a great, fun and intelligent person to be with.

A Company Built to Last

The funds we raised are dedicated to build a great company, that is here to stay. A company different from those backed by VC's institutional investment, that tend to window-dress for sale to the likes of Google, IBM, or - these days - Red Hat. VC money comes with the "VC clock" which asks for exit usually in a 3 year timeframe. With SOX, an IPO is usually not an option on a 3 years horizon, so the exit in 99% of the cases is an acquisition or a shut-down. Hence, virtually none of the recent venture investments has been made with the intention to build a company to last, but to build a company to sell.

Now, why does this matter to building an open source business? Isn't that self-serving?

Open Source is a Marathon, not a Sprint

I argue, you simply cannot build a true open source business in 3 years. Open source business is all about "users first, business second". All open source projects go through 3 phases:

1. The initative of one or a few individuals to bring an open source project to the starting line for community adoption
   
2. The community-driven rise of user adoption which still produces little or no revenues
   
3. The commercialization when the massive user adoption cries for services, add ons or other complements like db4o's non-commercial licensing alternative for embedded useage in non-GPL'd products
   

I think you cannot skip any of these three phases or parallelize them.

The benefit of keeping that order are drastically reduced sales and marketing cost, when a product is already widely adopted and generates inbound sales orders - a "product pull". This is the sustainable key competitive advantage of open source companies (and a huge barrier to entry for followers.)

Most VCs don't have this patience. In virtually all VC investments in OSS companies that I have seen, the management is forced to trigger step 3 right away. I see them mainly increasing their outbound sales and marketing spending to quickly ramp revenues for the sake of the VC's liquidity event - it becomes a "product push". I estimate that most open source companies are spending more than 50% on sales and marketing (rather than R&D), and I think a healthy threshold for OSS companies is 50% of budget spent on R&D.

These companies now ressemble much more conventional companies. They might have a 10-20% cost advantage, but that is something that any software company can match by squeezing operations.

Only true open source software companies have the power to move the possibility frontier by improving on both axes - quality and low cost - simultaneously, and they do it by spending on R&D, which results in user adoption and viral marketing, and ultimately yields in low cost, inbound sales.

What about the Post-VC Growth?

The Red Hat/JBoss deal has demonstrated, that after the VC exit you may be in a pole position to grow your business further with the power of a new parent, be it Red Hat, IBM, or Oracle. But what is the reasoning behind Red Hat paying $420M for a $16M revenue company? The reasoning is that Red Hat feels, they have the channel and bundeling power to "push" more product (rather than JBoss enduring 'organic' product adoption and awaiting the resulting "pull".)

Hence we're talking very conventional strategies as known by Microsoft, IBM and Oracle; there's no inherent differentiation of the business model any more.

Just Build a Great Product and Get Rewarded

So, we at db4objects, use the additional funds to increase our R&D spending to something like 2/3 of our total budget (an insanely high proportion by conventional standards).

And strangely, we see investment in good product coming back as marketing and sales mileage: Users and observers see our innovation speed, they spread the word, and they become more likely to adopt db4o, because we cover more use cases. And this ultimately results in more low cost, inbound sales.

A perfect world: You invest in good product, as defined by your users. And you see your order books fill magically.

Very different from the conventional, outbound sales model, which relies on vendor-lock-in and overpaid gung-hos and is, essentially, broken.