BusinessWeek: SaaS and Open Source are Good for Customers ... Unfortunately

"The Internet revolutionized the distribution of software—perhaps a bit too much. The Web brought a new, cheaper method for getting software into the hands of users, but in doing so may have killed one of the best models in Silicon Valley history."

"The challenge is you have to spend 50% to 100% plus of revenue in sales and marketing cost," he says. "You need this [limitless] amount of cash to forever feed the growth machine."

"On-demand software has turned out to be a brutal slog. Software sold "as a service" over the Web doesn't sell itself, even when it's cheaper and actually works."

She also takes a glancing blow at open source:

Summed up, her point is that software companies *loved* the old perpetual license business offering, because they got all their money up front, could charge separate yearly maintenance fees, and got an extra-special present in the form of an upgrade fee every few years as well. By contrast, when software is sold as a service, software revenues get spread across a longer time period, and maintenance and upgrades are bundled into the ongoing subscription fee. And this isn't nearly as fun for software companies.

I hear this kind of thing all the time from vendors. Somehow the shift away from lucrative up-front license fees is ... well, unfair. "We don't make as much. It's harder to get rich" goes the mantra of frustrated vendors. Left unrepresented in this bewailing about the injustices of the software business are customers. You know, the people who buy and (putatively) benefit from using the software.

The reality of this new world is vastly different than Lacy paints:

1. SaaS isn't necessarily any less expensive than packaged software. I looked at Salesforce pricing a couple of years ago, and it ran $900/year per user, or $2700 per user over three years. This is pretty comparable to packaged enterprise software. So the argument that somehow SaaS has financially ruined the software business reflects ignorance ofactual pricing in the marketplace.
2. One thing that *is* different is the pattern of cash flows. Packaged software attains a significant portion of its lifetime revenue cash flows at time of sale, so there is cash to feed an extensive marketing and saleseffort (Lacy refers to this as the days when "swaggering, elephant hunter-style salesmen would drive up in their gleaming BMWs to close massive orders in the waning days of the quarter"). SaaS receives its revenues much more evenly over the course of the contract, so there is less money received quickly to fund those elephant hunters. In this respect, SaaS resembles a subscription business rather than a product business.
3. The reason why it's harder to grow SaaS reflects the reality of subscription-type businesses. These companies *always* have customer acquisition costs which must be amortized across the life of the contract and which restrict the amount of cash available to invest in growth. The challenge for subscription businesses is to balance growth against cash flow.
4. Just because it's harder for vendors doesn't make it wrong, and, in fact, there's an argument to be made that customers are finally getting their due. Just because the packaged software business was a pleasure cruise doesn't mean that the party would last forever. Most businesses are a brutal slog. Software isn't as lucrative a business as it once was. So what? Forcing software vendors to deliver better value is good for customers.

Turning to an unexplored, but extremely critical part of Lacy's piece, let's look at the issue she barely mentions: the level of customer hand-holding. During the heyday of packaged software, customers (at least those customers promising very large software purchases) could expect unlimited attention from sales personnel, including educational seminars, vendor CTO visits, free proof of concept implementations, etc.

Some of that still goes on with SaaS companies, because they can amortize the cost across several years of subscription fees. However, they ration the effort, given the realities of customer acquisition cost amortization. That's why it's hard for SaaS to grow as quickly as packaged software companies could.

Turning to open source companies, however, it's completely different. They typically sell support contracts, so it's something akin to a subscription. But customers usually only buy *after* the system goes into production, long after any selling/customer hand-holding effort is performed. And only a small percentage of users who go into production end up purchasing support.

Therefore, open source companies do very little traditional software sales activities. There is no free hand-holding. I heard the Chairman of one open source company state that their sales people are told to discard any sales lead if it takes more than 30 seconds to move it along to the next step. The implications of this are quite clear. If you're a potential user, you need to be able to self-sell, i.e., choose, evaluate, and pilot your own software decisions. When it comes to open source, the customer faces the slog, not the vendor. If you're a self-starter, so to speak, this world is great -- high quality software, low prices. If you're a user that likes a lot of vendor attention, you'll just pay in a different way. Instead of funding a sales rep's BMW via the license fee, you'll pay the vendor's consulting group or a system integrator to help you in the decision process.

While many in the Valley may lament this change, and pine for the good old days, I think I prefer the new way. It maps value received to costs more directly, and gives the vendor and service providers more incentive for customer satisfaction.