Accuse me all you want for having an alarming title like the above, but I’m sticking to my guns. A lot is wrong with the venture model in general, but the malaise that has afflicted IT VC has reached an incurable situation. I’ll summarize what I see happening and then explain it all a little better.
This is the new reality:
- Software is no longer a capital intensive industry (this has many important causes and corollaries which I’ll get to shortly).
- More often than not, close-to-free guerrilla marketing is about as effective as massive traditional ad spend.
- A successful software company doesn’t necessarily equate to 100X or 1000X returns on the initial investment.
- IT is now a fact of life. Dare I say it, a commodity. A cool new IT product evokes about the same reaction that an interesting freebie found in a box of cereal does. Not because they bring the same value, but simply because we’re all so used to IT now. There are no unreasonable premiums to be had. Bye bye $100,000 static HTML websites.
- Traditional channels of distribution - which were incredibly expensive to build – are out of the picture.
Before I go on, let me just tell you… if you look hard enough you can find examples that contradict one or more of the statements above. Take the virtualization industry. How can you really distribute a hypervisor and expect people to implement it in a large corporate environment without murderously expensive high-touch sales? Or without fancy implementation and consulting firms delivering your solution? You would be right in pointing out such examples, but only partly. And only temporarily.
The fact of the matter is that in the vast majority of enterprise software sales and implementations situations where you aren’t the deeply ‘embedded’ multi-billion dollar corporation that already owns the account, a high touch model such as the one above no longer works. And by not working, I mean it seldom makes money. There is so much desperate or free competition, so many ‘built in for nothing extra’ features and such a long sales cycle that if the startup vendor walks away with only a $50,000 loss and a nice looking success story with a customer quote, we’ll all smile and call it a success.
So once again, why is this happening? Let’s take a closer look.
Software is no longer a capital intensive industry. Traditionally, the costs associated with software development have been things like labour, licenses, rent and systems. That’s pretty much it in a nutshell. On the labour side, there are two primary variables that have changed the game considerably. The first, of course is the basic per-hour cost. With the success of the offshore model you can get services cheaper today than ever before. The second, which is even more important, is productivity. With the free availability of every kind of library imaginable, with billions of pre-built ‘cut-and-paste’ code samples, with an online collection of how-to resources that’s growing every second and with rapid advances in development tools such as IDEs, developers are more productive today than ever before. It is possible for ‘one guy’ to build a meaningful online service that will be used by millions of people. For the not-so-essential functionality, simply outsource it for a few dollars an hour.
Then there are licenses and tools. Many years ago, when I built my first startup, this was quite a significant expense. The best IDE available at the time was probably JBuilder. I remember we paid about $1,000 per seat for each of our 20-25 developers. The source repository we were using was Microsoft VSS, which was another few hundred dollars per seat. And the QA team tested with webload, which I distinctly remember was an almost $100,000 license for the level of functionality we needed. Keep in mind, we’re talking decade-old-dollars here. For comparative costs in 2009 money, multiply everything by 2 or 3.
Today, the best Java IDE is Eclipse. It is completely free and far more functional. There are hundreds of free plugin modules developed as open source projects that extend its functionality. The most popular source repository is probably CVS, which is also open source and completely free. There are dozens of free load testing tools which may not be as good as webload, but they get the job done quite effectively. So as not to completely have you fall asleep, I’m just giving you a few examples. The reality is there are many more where those came from.
Finally, the hardware. I don’t even have to tell you that PCs have gotten cheaper. Duh. But it’s really beyond cheap, at this point. They’re almost free. And frankly, while I think the PC industry has essentially written its own epitaph by substituting a $1,000 PC ASP (Average selling price) with $250 or so for netbooks that can plug into flat screen TVs, the bottom line is that software startups really don’t have to worry about this expense like they once used to. A QA lab with 20 machines doesn’t cost more than $7 or $8k. Comfortably within the limits of most credit cards. AND, with Virtualization you can do things on these 20 systems that it would have taken a very large facility to accomplish back in 1999 or 2000.
Before we move on from the capital intensiveness argument, a corollary to be noted is that capital is, traditionally, a barrier to entry. So if capital requirements in an area are very low, one guy can come up with something that upsets an empire. Rather than being an interesting rarity, this sort of disruptiveness increasingly becomes the rule. VCs, then, should be uncomfortable. Their money is no longer a guarantee of any kind of protection.
To be continued… When we come back we’ll look at some other reasons, including why zero cost guerrilla marketing is becoming increasingly effective and why rational return on IT investments irreparably break the VC model.