Original URL: http://www.theregister.co.uk/2007/05/01/the_new_software_industry/

'It'll be ugly when half the software industry goes away' - pundit

If by 'ugly,' you mean great

By Ashlee Vance

Posted in Applications, 1st May 2007 19:36 GMT

Analysis An imploding software industry could be just what customers need. You're going to end up with cheap, specialized software and all kinds of wonderful services to support the code.

We're pretty sure that's what executives, researchers and developers told us yesterday, during a Carnegie Mellon West sponsored event at Microsoft's Silicon Valley headquarters.

Frankly, the "New Software Industry" event failed to cover much "new" ground. Former Oracle President Ray Lane kicked off the gig by promising continued consolidation in the software industry. Then, a cavalcade of other speakers documented software companies' push to replace licensing revenue with services revenue. After that, a couple of speakers hyped up open source as the dominant force in the software game.

Anyone with a web browser and even a passing interest in software has been made aware of these three trends. Oracle, Cisco, EMC, HP and IBM all do their part to stress the consolidation theme by buying as many companies as possible each week. The same companies – along with everyone else – have flogged their services too. And no one is safe from the rhetoric around the magical "army" of open source coders who release product that undercuts proprietary software.

So, we're going to try and focus on the money aspect of all this. We hear that customers care about their bottom lines.

"The software business, I think, is at a crossroads," Lane said. Being even more direct, he added, "Software prices are going to fall rapidly."

Lane sees "a large no man's land" of old line software makers trapped between Microsoft, Oracle and SAP, which consume 75 per cent of the available revenue, and start-ups such as Salesforce.com that carry no baggage. Such software makers may seem strong with, say, $300m or so in market capitalization, but they're really just waiting in line at the slaughterhouse. They don't have the might to compete against the big three or the cost structures capable of defeating new players.

"Will Oracle continue to put pressure on BEA? You bet. They will continue to outspend them all day long," Lane said.

"It will be pretty ugly when half of the software industry goes away, (but) it needs to happen."

A couple of financial types backed Lane's sentiments.

Bill Burnham, of Inductive Capital, described the aging middleware crowd as "the walking dead."

He characterized Tibco and BEA as "decent" stocks to short - a method of investing where shareholders bank on the share price of a company falling. Ann Winblad, of venture capital firm Hummer Winblad, added to the bloodbath by picking on CA.

"I would certainly short Computer Associates," she said. "They are in a lot of old businesses."

BEA, in particular, has taken similar shots for years. Investors punished its stock in 2001, after your reporter confirmed that Sun Microsystems would give away its application server. Few customers, however, actually moved away from BEA in the end.

The hottest software company on the planet - VMware - also seems to go against the doom and gloom scenario presented by the pundits. Close to a decade after it started, VMware continues to show 100 per cent year-on-year revenue growth and has an IPO on tap. It makes far more money running multiple copies of Linux on servers than the open source, services darling of the "New Software Industry" analysts - Red Hat.

Enduring Reality

But defending per processor-priced, proprietary software makers does seem a futile battle.

The big three all have so-called "on-demand" strategies underway for obvious reasons. New license revenue has stalled. Making matters worse, open source companies dish out new, good enough software for low-end tasks at a steady pace. This leaves services and maintenance programs as the only real ways to make new money.

Oracle, according to former Oracle On-Demand chief Timothy Chou, "would be lucky" to pay $60 per user to ship its software as service due to its infrastructure costs. That compares to $7 per user for Salesforce.com and 70 cents per user for "mother Google."

So, the big fellas have an enormous challenge ahead as they try to compete against fresher rivals. Oracle, SAP, Microsoft and those in the no man's land have decades of baggage to support that adds a ton of cost to their "software as a service" efforts, Chou said.

Such figures back up the notion of a no man's land collapse. The Big Three can afford the software as a service transition thanks to their ample cash stockpiles, while smaller companies will have to give up or be swallowed.

Now that you know how the middleware crowd will be rocked, let's get on with that pricing panacea that we promised.

If you're going to "software as a service," you better have good services to back up the shift. So far, that hasn't really played out.

You all know the horror stories about an IBM consulting team visiting your data center for a one-off engagement only to end up pitching tents and setting up mini-fridges for what turns into a multi-year stay. Three years and $10m later, they tell you that DB2 might have been the better choice than Oracle after all. Time for another software roll out.

IBM's own software researcher Paul Maglio thinks very serious work needs to be done on the services front for the good of vendors and, quite frankly, for the good of freedom, justice and the American way.

"We have to focus on services and services innovation," he said.

IBM, for example, has "grown its services business like crazy," but the services unit has lower gross margins than the even the ultra-competitive corporate hardware business.

"The profit in the services business does not grow proportionally," Maglio said.

Rivals such as Oracle, HP and Microsoft face the same issue, and that's at least partially their fault.

Maglio argued that the US has failed to spend real money looking at innovation in the services field.

"We have no Gordon Moore of services.

"There is no guiding star. There is nothing telling me to invest in people in this situation and to invest in technology in this situation."

IBM has tasked Maglio and others with trying to improve the communications between the vendor and customer in the hopes that something more concrete can come out of services engagements. Such research should go well beyond the same-old, same-old case studies. Maglio wants to develop a rich methodology for understanding how to make Big Blue consultants useful in a broad set of situations.

How to get decently rich from software

While IBM tweaks the services angle, a host of companies continue to focus on the actual software piece of software as a service.

If, as the pundits say, large, platform-style software companies are dead, then what's left?

Well, the consensus seemed to be that small, very focused software companies will thrive in the next few years.

For example, Winblad pointed to one of her babies - Mule Source. The start-up makes little more than an open source Enterprise Services Bus that goes up against code from Tibco and BEA.

"They are growing very rapidly," Winblad said, as you might expect.

Similarly, Chou flagged Flock IT - a poultry management package.

Such specialization will only increase as the costs to start a new software company and distribute the code drop thanks to better bandwidth and cheaper hardware.

"You must realize that specialization is going to be a huge influencer," Chou said.

In the hardware game, companies must still spend billions of dollars on research and development to find the next big thing. That's just not true in the software realm.

"In the world of software, that is totally reversed," Chou said. "The research lab is the research lab of the internet."

We'll close with Winblad's thoughts on the software game, since she seemed to have by far the best grasp of the topic at hand.

The VC championed data center management and analytics software as two of the more interesting future markets.

There's plenty of room for another VMware, according to Winblad, because the virtualization game is so young. "It is still very new technology to the customer base," she said. "Even VMware and Microsoft with VirtualServer are very small compared to the total market."

She adds that companies such as Omniture seem on the right track with their data analysis software.

And with that, here's hoping a cruel editor doesn't force your reporter to listen to "software as a service" torture again anytime soon. ®