I posted recently about how the concepts enterprise software market, software as a service (SaaS), public cloud computing and open source are so often mixed up and treated as though they are different IT investment opportunities. When the post appeared on seekingalpha.com (see link below) -- as some of my posts do -- Kevin Dobbs posted a reply saying that SaaS was different. I repeat his comment and my reply below on how I think they are linked.
But in thinking about my reply to Kevin it dawned on me that my analysis that SaaS is only 5% or so of enterprise software spend is misleading. Given that SaaS subscription fees also cover a CIO's personnel costs and hardware costs as well as software license and maintenance (content update) costs, the better way to analyze it is to look at SaaS as a percentage of overall IT spend--which could be somewhere north of $10 trillion depending on what you include.
Looked at this way, SaaS so far represents much less than 1% of IT spending. But I still believe it will grow to dominate. To dredge up the old utility analogy, there's still too many CIOs building and running their own dynamos.
George Westinghouse, where are you?
-- Dennis Byron
(no financial interest in companies mentioned)
Comment received September 24:
Dennis, you are replaying the common wisdom that most of the old line analyst firms are touting. One point of clarification is that Software-as-a-Service is actually a different business model than the license software business model. SaaS companies deliver products that are more like consumer-based software (think Facebook), and have a different approach to sales, marketing, services and support. The best in class firms are much different than Oracle, SAP, IBM, HP and others. Customer buying preferences are trending towards a subscription model and that's why you are seeing all of the technology master brands moving in this direction.
Reply posted on SeekingAlpha.com
Kevin, thanks for your comment even though I am about to disagree with you totally.
First of all, I am not replaying the "old-line analyst" opinion. As the picture shows, I am the old-line analyst. I have been saying this for 20 years. (However, I make a subtle distinction: I am an IT market researcher, not an analyst in the sense I think you mean: I have not been the type of analyst that gives users advice since I left Datapro.)
When I started in IT marketing 40 years ago, SaaS was called timesharing, then service bureaus, went through a few more iterations and by about 1999 was called application service provision. I am not proud to say that I participated in coming up with that term. We had the numbers right at IDC back then but basically too much hype got generated over what we clearly said were low numbers as a percent of total market spend (as I mentioned in my article above).
There is nothing new about SaaS as a business model or otherwise although some market researchers looking for differences (which is good; that's their job) parse things -- like multitenancy and whether the software is exclusively available as a service or also available under a perpetual license -- to see if there are differences. But in the end, the difference is so minimal as to have little effect on market dynamics. The user is going to pay x over seven years for y functionality. It doesn't tend to matter a lot whether the user pays 5% of x on day one -- the licence price -- and then the remaining 95% monthly over the next 84 months -- the subscription maintenance to keep tax tables, etc. up to date, the hardware costs, and most significantly the personnel costs. Or whether the user just spreads the whole 100% out over the 84 months.
I don't get the "SaaS products are like Facebook" thing at all. I know you are parroting the Marc Benioff party line but as I went into in detail in February 2010 (see byrondennis.typepad.co...), that's just good PR. I complimented CRM for it.
I am not sure which software suppliers you mean by "best in class" firms but as I explained last year (see byrondennis.typepad.co...) the only difference between IBM, Oracle, etc. and the so-called SaaS pureplays is that the latter are not really pureplays. All began trying to sell their software via a perpetual license and failed. And conversely, Oracle may already be second only to CRM in SaaS revenue (and if not, it will be within a year or so to be followed, depending on corporate strategies by SAP, Microsoft, Intuit and possibly -- although it's a special case because of its mainframe business -- IBM).
All the indisputable market research issues aside, as I said in my article above,
"I believed then and still believe that SaaS will again become a dominant enterprise software delivery method, just as it was when I entered the business 40 years ago."
It's just that that trend has nothing to do with a different business model, Facebook, different sales or marketing techniques, technologies, or pureplay vs. stack.Thanks again for the comment