Thursday, August 16, 2007

VC greed and the open source cause

VCs fund companies so they can get 10x or 100x of their money back. They don’t want to own a company, they want to flip it. This, of course, determines the growth trajectories and exit strategies of the firms they fund, including open source startups.

Two weeks ago, publisher Tim O’Reilly wrote:

FOSS companies have essentially the same business model as Microsoft, except that they've largely given up on revenue from the consumer market and instead target the corporate market. (So let's say they have the same business model as Oracle, just to be more precise.) And while OSS is far superior to proprietary software as a development model, and has some advantages in lower marketing costs, a close study of any FOSS companies shows that they have no demonstrated business model advantages over Microsoft's or Oracle. …

I will predict that virtually every open source company (including Red Hat) will eventually be acquired by a big proprietary software company.

The real change to proprietary software is going to come from the Web 2.0 challenge.
O’Reilly used to make money publishing Linux books and now he makes money publishing books on Web 2.0, a term that he invented. Hmmm. Still, he has proven a clever trend-spotter in the past.

Yesterday, my friend Matt Asay tried to rebut the O’Reilly claim. Asay was formerly a GM at Lineo (an embedded Linux company) before they cratered, then director of Novell’s Linux business office, and now a VP at Alfresco, an open source content management company.Matt claims:
I actually believe that open source, not proprietary software, is the natural state of the industry, and that Tim's proprietary world is anomalous. …

Suddenly, the license matters more, not less, because it is the license that ensures the conversation focuses on the right topic — service — rather than on inane jabberings that only vendors care about. You know, like intellectual property. …

I look forward to buying Microsoft. I've got a great job ready for Steve a product evangelist.
I’m sure Matt is neither smoking something nor drinking anything stronger than Coca-Cola, so I guess he just has a fertile imagination.

Those VC-backed startups need an exit strategy to make the VCs happy, and (absent something silly like an ESOP) the choices are limited to IPO or be acquired. Yes Red Hat and a few other OSS companies IPO’d in the red-hot 90s, but today that’s very unlikely for any software company. There are noises that MySQL is hoping to IPO, but how many other OSS companies are there with a similar installed base?

Instead, the normal path is to be acquired. Embedded database company Sleepycat was bought in Feb. 2006 by Oracle, and yesterday virtualization company XenSource sold itself to Citrix, a close ally of Microsoft.

VCs sell their OSS startups to proprietary software companies because (as Willie Sutton didn’t say) that’s where the money is. So as long as impatient investors are looking for liquidity — and the IPO market remains closed to all but a handful of companies — OSS companies will be gobbled up by proprietary software companies.

That’s not to say that Matt is wrong on the other stuff. The software industry is likely to shift from selling licenses to selling service, as product categories mature and users revolt against upgrades with unneeded features. This was a “trend” when I was an ISV 20 years ago: companies keep paying for support and assurance, not for creeping featuritis.


John Newton said...

It's a matter of how you organize your company. At Alfresco, we are well structured to ultimately go public. If you do not plan to do so, then it is difficult to do an IPO.

I have been in closed source companies, including founding Documentum, and now an open source company. So far, our experience has been that open source grows faster than closed source, and with the self-seeding, self-trialing model, open source is a lot less painful.

My experience is that we don't need a big company acquisition to be successful or to fulfill our investors objectives.

Matt Asay said...

I agree with John here (small wonder, no doubt :-). It is true that the primary exit for most companies - open source or proprietary - is to be acquired. And given that most of the existing market is proprietary software today, it's reasonable to say that odds are in favor of proprietary companies buying open-source companies.

Still, I believe that there will be plenty of good open-source companies who manage to go public, rather than be acquired. Given Alfresco's current traction, I firmly believe we'll be one of the first.

That said, I actually think Tim wasn't making nearly as placid an argument as yours (or mine). I think he was implying that open-source companies lack the horsepower; that the future (and present) belongs to proprietary. He doesn't paint this in nefarious hues. He's talking about kind, loving proprietary Web 2.0 vendors and such.

I still think he's wrong. Tim has imagined the future correctly on several occasions, but he's maybe a little less adept at remembering the past. The past goes on for a long, long time. SaaS, for example, may be the future, but we're going to be dealing with on-premise server software for decades, even if SaaS completely takes over tomorrow. There are still green screens all over the enterprise.

I believe "packaged" open-source software has a very healthy future, standalone and as acquired by open-source companies or proprietary vendors.

Joel West said...

It’s nice to get two officers of the same startup to read my blog. I guess I should name drop more often. :-)

My post wasn’t actually about Alfresco, but about a blogger whose day job is at Alfresco. Although the odds are long for any software to go public, I’ll certainly defer to the chairman of Alfresco as to how well they are preparing for a possible IPO.

However, other entrepreneurs have been overly optimistic in the past. My point was that exit strategies are often determined by the cycles (or long-term trends) of the capital market rather than a given company’s business model. My mentor Charlie Jackson structured Silicon Beach Software for IPO, but then the markets turned sour on startups and selling to Aldus was his best option.

Perhaps Alfesco will go public. Even if it does, that doesn’t prove that IPO is a realistic exit strategy for any significant number of open source companies. We don’t see a lot of airlines or car makers IPO’ing either — once an industry starts to mature, it’s hard to grow to a scale comparable to the incumbents. So a few companies will make it, but I’d be shocked if we had as many software IPOs in the 2000s as in the 1980s.

Matt knows and pays more attention to Tim O’Reilly than I do. If he thinks that Tim is making a stronger claim about the future of OSS companies than I had assumed, then my beliefs are somewhere in the middle between Tim & Matt. I think proprietary software companies will sell services, open source software companies will sell services, and in the long run it will be hard to tell the two apart.

However, in the year 2025 (if man is still alive), the majority of big US software companies will be companies will be descended from today’s big companies. The only significant numbers of new big software companies will be those from India, China and some other country we’re not even thinking about right now. (I’d say Russia, but there’s little evidence that the ruling class will embrace capitalism between now and then.)

Maria said...

I’ve read that there are high quality contributions to OSS projects from other countries, such as Eastern European countries (which has an explosive percentage of contributions to OSS projects according to I can see why open source companies have an edge on proprietary source companies as a service provider; they can accelerate innovation and global software delivery at a reduced development cost.

In this cycle of the capital market, is it more difficult for a company to IPO if it sells services rather than software licenses? Or are VCs more interested in growth projections?