Software, Interrupted

IDC recently published its annual report on software industry market share, which ranks software companies by revenue. According to IDC, despite Oracle's string of large acquisitions, including its pending acquisition of Sun, it will remain in third place behind Microsoft in first and IBM in second.

I received some interesting data points from the IBM PR team:

• IBM's software revenue in 2008 totaled $22 billion.
• Software represents a whopping 40 percent of IBM's overall profit, and 20 percent of its revenue.
• In 2008, more than 90 percent of IBM's segment profit was from software, services, and financing.
• IBM has acquired 81 software companies since 2003 and more than 100 software companies in the past decade (including Cognos, Filenet, Telelogic, Micromuse, and MRO Software).
• The R&D focus at IBM has shifted more toward software and services. More than 70 percent of the U.S. patents IBM received in 2008 (IBM's 16th straight year of patent leadership), were for software and services.
• IBM has been driving a shift to higher-profit segments (versus the low-margin commodity parts, which is why it got out of the PC, hard drive and DRAM businesses.)
• IBM is also getting greater margins from creating offerings that exploit the blurring of software and services, such as cloud computing and SOA.

A more difficult but possibly more interesting metric would be to understand the levels of adoption--that is, servers, users, etc., versus revenue--to get a better picture of the market. Of course, revenue and profit are all that really matters, but I'd like to know if IBM makes more money than Oracle on a potentially smaller installed base.

The big mystery is who else can Oracle buy in order to overtake IBM? It seems like industry consolidation has taken out most of the obvious candidates. Maybe this is where we'll start seeing Oracle acquire SaaS providers in order to gain fast-growing market share?

Follow me on Twitter @daveofdoom

In July, SAP instituted a new policy that forced users into paying a higher support cost with no clearly defined additional value for the higher cost.

ZDNet's Dennis Howlett interviewed Alan Bowling, chairman SAP user group UK and Ray Wang, VP at Forrester research about the price hikes and what they really mean for users.

The short version: SAP users in the UK and Ireland will not accept the next round of price rises for support unless the company has "clearly delivered improved total cost of ownership" for the businesses that use its software.

Software licensing is coming under more pressure in the difficult economy and support and maintenance fees are clearly important. SAP needs to realize that every piece of software is replaceable--switching costs may be high, but being held ransom is far worse.

SAP has announced that it plans to transition all customers to its new Enterprise Support offering starting January 1st, 2009. There's nothing like being stuck on a giant system like SAP with no choice but to pay whatever ransom they ask for.

This reminds me of the time that Hoboken, NJ parking garage didn't want to pay for a 20% license increase and all the cars got trapped.

According to the press release, SAP will incrementally raise support costs up to 22% by 2012: "For example, subject to specific contractual limitations, a typical customer paying 17 percent of maintenance base for SAP Standard Support in 2008 will pay a rate of 18.3 percent of maintenance base for SAP Enterprise Support in 2009."

While this feels like sneaky attempt to raise prices (see previous post on Oracle) at least they are not raising license fees. There is value in support, though the increased cost doesn't exactly explain what you get for the additional cash.

Via Irregular Enterprise (ZDNet)

This is the second massive SAP mishap this year. First was Waste Management's lawsuit for $100 million, and now Levi's couldn't ship product for an entire week.

Gavin at The Reg caught this story last week but I just saw it on The Guardian. Where's the link, Guardian?

From guardian.co.uk

After a sales and profitability comeback in 2007, the jeans maker was confident going into a tough 2008 that, despite a global slowdown, progress would continue.

But what Levi's was unprepared for was the fact that the company's IT operation could not get its multiyear, multimillion dollar investment in SAP's business software to work optimally.

The situation became so bad that early in the second quarter, the system was shut down completely for a week. On a conference call a few days ago to announce a 98 percent drop in second-quarter profitability, CEO John Anderson said the computer disaster was a "substantial" factor in a 19 percent decline in U.S. sales and the subsequent profit meltdown.

SAP's Henning Kagermann thinks that big companies are too conservative (afraid?) to adopt SaaS.

But while it is the sales people, numbers crunchers and purchasing pros who enter information into these systems, traditional business software isn't designed with these workers in mind. It's designed for these workers' bosses, who need a single place to track what their departments and employees are up to.

With every other company trying to increase productivity and bring the consumer/collaborative angle into their software it's interesting that SAP chooses to remain archaic and conservative. Only time will tell if that's the right choice.

One company that continues to baffle me with their non-presence in SaaS is SAP. Nearly two years ago Matt Asay and I sat shocked on a CSFB panel just after Shai Agassi proclaimed that SAP had spent over $1 billion on BBD before it even got out the door.

Over on the Enterprise Anti-matter blog Josh Greenbaum finally finds out why SAP BBD is so late and potentially DOA--a fundamental design flaw in the system, along with what seems to be a misunderstanding of building large-scale applications.

The main issue regarding the delay of BBD has to do with operationalizing the on-demand model in a cost-effective way. This translates to a realization that the existing release of BBD, now in use by some 150 customers, can't be scaled up to handle thousands of customers in a cost-effective manner. This is a major operational problem to be sure: SAP can ill afford to ramp up to a massive deployment of BBD if it's not cost-effective to do so. Klaey mentioned that these operational issues seem to have been dealt with in the next release, which is slated to come on line in the "next few months."

I am on the "Integration Behind the Fire Wall - Take II" panel tomorrow at 1:30pm at the OpSource SaaS Summit here in SF.

It should be a good one with panelists from SAP, Cast Iron and others. We'll be talking about the enormous burden of integrating enterprise apps with SaaS.

Whoever mentions this blog post will get a squeezy Mule until I run out of the few I can carry.

One of the things that I have noted in the past is the fact that ongoing consolidation limits the choices for IT buyers. There will eventually be a backlash against these behemoths and we'll see more open source and SaaS alternatives take the place of these giants.

Today's WSJ notes As Software Firms Merge, Synergy Is Elusive:

The issue of what customers experience after a big tech merger is once again coming to the fore as the software industry undergoes its latest wave of consolidation. International Business Machines Corp. last week plunked down $5 billion to buy software maker Cognos Inc., while Germany's SAP AG recently agreed to purchase France's software maker Business Objects SA for $6.8 billion. Meanwhile, Oracle last month made a failed bid for software maker BEA Systems Inc., and has in recent years also purchased software makers Siebel Systems Inc., Retek Inc. and Hyperion Solutions Corp.

Obviously, you always hope for the best but when you make a commitment to software that you don't control it should keep you up at night. You are at the mercy of the market and corporate whims.

As much as I like open source, there is something to be said for not having to install or maintain applications. But that doesn't mean it's smart or realistic to move all the applications in your enterprise to on-demand delivery.

As Gary Rivlin writes in today's NY Times "few software companies make the move to platform status" let alone the ones that have no footprint on the desktop. This is something I noted back in September after the Dreamforce event which reinforced my belief that enterprise software isn't going anywhere--at least not infrastructure software.

In the case of infrastructure (like networking and integration), and desktop environments (like Windows) it's hard to get excited about doing everything through a browser. Not that it?s a bad idea, but as I learned on my trip to Japan last week it's not entirely feasible at this point to do everything via the internet. Rivlin writes:

And yet for Benioff, the company's chief executive, that is not enough. He wants to turn Salesforce into a platform like Microsoft's Windows operating system, a product so popular that it is the foundation for a veritable ecosystem of software developers.

If you at the on-demand subscription offerings from Microsoft and SAP, both mediocre by comparison to Salesforce.com or SugarCRM (also available open source) the main advantage they have is that the existing user base is tied to a set of desktop applications which reinforce the desktop computing paradigm.

... Read more