Perpetual licensing is dead. Long live SaaS!

That was not the official theme of Software 2007, it was actually "powered by innovation".  Nonetheless, there was a lot of talk about on-demand software, software as a service (SaaS) and "Enterprise 2.0".  Perhaps because these are the key drivers of innovation in the software industry, and they are as much about the market and economics as technology.

I thought I should attend the "Venture Capital Panel" breakout session, since everyone has been telling me that VCs don't invest in enterprise software any more. 

The panel was moderated by Dan Dorosin (Fenwick & West) and featured Bob Lisbonne (Matrix Partners), Ravi Mohan (Shasta Ventures), and Ted Schlein (Kleiner Perkins Caufield & Byers).

Some key points:

• Look for "white space" above, below, and to the sides of existing applications, not squeezed between them.
• Look for models borrowed from Web 2.0, where adoption is driven at the user level, rather than dictated from corporate headquarters.
• Bob Lisbonne quoted one of his mentors: " There are only two ways to make money in software: bundling or unbundling."
• The big ERP vendors "own" the CIOs office. Grow virally by selling to buyers who have a highly specific pain points, their own budgets, and minimal control from corporate IT.
• Show investors that your value proposition is so compelling that you are able to sell to big companies without a big sales force.
• Have a great sales partners.
• Consider a mix of software and hardware (e.g. appliances).
• The days of big upfront license fees are gone - make sure you have an on-demand option.
• However these deferred revenues increase the amount of capital required.  This cash flow problem can partially (or totally) offset the trend for lower costs of software, bandwidth, and infrastructure.
• Estimates of the amount of capital required was the subject of some debate. Estimates ranged from $25 million to $100 million. The discussion seemed to settle around $40 million.
• Many SaaS companies are negotiating one to two years of subscriptions up front to help ease cash flow.
• On the other hand, there are examples of companies that have bootstrapped to $6 million to $10 million in revenue.

Bottom Line: The approach to the market is as important as the technology because it is very difficult to compete with the big guys.  Do some smart product marketing and fight the battles you can win. Maybe you'll get some investment.

Comments

[Justin]

$25-100 million in capital to start a software as service business? That is insane. You can get a ton done with 2-3 great developers, 1-2 racks of servers, and maybe 1-2 biz people.
Venture capitalists inherently believe that businesses need huge amounts of capital to grow, because that is what they bring to the table. For a software company it's just not the case anymore unless you're doing something wrong. IMHO.

[Paul]

Yep, their estimates surprised me too.

Bear in mind they are talking about a "typical" enterprise software VC deal. They are not just talking about starting the business, they are also including the cash required to scale the business to $50-$100 million in revenues.

One of the issues for enterprise software is the cost of sales. Viral sales & marketing is not as applicable to enterprise software as to consumer software because of group decision making. For example, on-demand players such as Arena Solutions and Kinaxis often find themselves in a formal beauty contest against traditional vendors.

I have seen estimates of cost of sales for traditional enterprise software of $70,000-$100,000 per sale. It would be great if anyone has updated figures.

Also, the traditional view is to expand functionality (to grow same customer revenue) while keeping the core technology up to date. That requires more cash, more developers and perhaps a more formal product management process.

The dominant costs will be people related as the company gets bigger.

So I think their point is that you can bootstrap to $6-$10 million in revenue, and have a nice little business. But (in general) you need a lot more capital to grow and scale to the $50-$100 million they are expecting to make the VC model work.

I would love to hear more points of view on this issue. How much capital do you really need for massive scaling in the enterprise software world?