CFOs start to see the benefits of open source

CFO Magazine is running a great story about the cost savings available from open-source software. This is a topic that you'll hear open-source vendors crow about, but it's somewhat rare to actually get a CFO on the record about her benefits from open source, so it's notable.

Open source is gaining at proprietary software's expense.

(Credit: Gartner)

Recent Gartner research suggests that over 27 percent of enterprises will deploy open-source software in 2009. (Note: the remaining 63 percent will, too, but Gartner must have asked the CIO, and the CIO is the last to know.) That's up from 25 percent in 2008, while the share of proprietary software deployed will actually go down.

That is perhaps the scariest data point for proprietary vendors who have to grapple with former safe havens like InterContinental Hotel Group, which CFO Magazine notes is adopting Red Hat, Alfresco, SugarCRM, and MySQL, among other open-source products, for its mission-critical systems going forward.

Against this backdrop CFO Magazine calls out the benefits of open source in a tight economy:

One of the initial raps against [open source] was that, while the idea of free and continuously modified software had a certain appeal, it also inspired a certain terror; what business would hitch its technological star to software that was pulled off the Web and unsupported by a major vendor? Who knew what lurked in the code, or how easily that code might be cracked into?

Today, the recession and its attendant impact on IT budgets have prompted companies to live with a certain level of anxiety. And, as well, years of experience by those on the cutting edge have shown that many applications within the [open-source] world may now be ready for prime time. Vendors do in fact play a role in supporting [open source], and while their fees have been rising, overall cost of ownership is still substantially lower; often that vendor support feels more like a security blanket than a shakedown.

This isn't a Payless Shoe Store commercial, either ("You could pay more, but why?"), promoting goods of generally less quality. Open source has hit its stride, and often the open-source competition is actually better for enterprise requirements than the proprietary alternative. For example, if an enterprise is running Web applications, it would be daft to not at least consider using the leading Web database: MySQL.

Better software, lower price. What's not to love?

If you're a CFO, there's much to love in open source. If you compete with open source, well, perhaps you won't be so enamored. Sorry about that. Competition has returned to software.

Disclosure: I am an Alfresco employee, and an adviser to SugarCRM.

[ian.waring]

In a recession, it's the CFO calling the shots more often than not. Further traction, here it comes...

[ITRebel]

Matt,

As usual, I appreciate the great information. As usual, I will offer a counterexample though. Here is a very good example of why a proprietary product called Matlab has much more appeal than an open source language called R in my field of advanced analytics. This benchmark comparison appeared in today's R-help post for the R language. I am quoting directly from the summary of the benchmark comparison in the post to today's R-help:

"The machine used for testing was an i386 core 2 duo machine at 2.2
GHz, running OS X `Tiger'. The R was 2.8.0 and the matlab was 2008a.

Here's the performance (time taken in seconds):

Matlab R Times
Vector version 0.0446 0.067 1.5x
For loop version 0.0992 42.209 425.5x

So the R is 1.5x costlier for the vector version and 425.5x costlier
with matlab.

I wonder what we're doing wrong!"

After this post, rhere was much discussion about how R users need to find ways of programming that do not use For Loops and instead use Vectors, but good luck! For any significant machine learning method, a substantial amount of For Looping is absolutely necessary. Obviously the fact that R is free does not matter a bit when one wants a solution to a problem. Many of these advanced machine learning methods that require For Loops take many hours to run anyway in Matlab. An open source and free software application like R is actually much, much, much more costly in this example when you multiply it across the number of users in an organization and the number of times that you use the program.

Matlab has gotten to this point of such awe inspiring speed through proprietary and closed innovation. This is a staggering example of the advantage of proprietary software over the Freebie stuff.