Why pay SAP for software you don't use?

I took a lot of heat last week from SAP and its friends over my posting about The scandal of SAP's idle CRM seats. Seems like I struck a raw nerve when I wrote:

"The dirty secret why SAP had to launch an on-demand CRM offering is that its on-premises product is too hard to implement ... as many as two thirds of the CRM licences SAP has sold are lying idle, according to one estimate."

First of all I got an email from SAP's PR company stating that my If I were one of those customers, I'd be really fuming by now"accusations and conclusions" were based on "erroneous information" in the source article at SearchSAP.com that I had linked to. I'm sorry to see that SearchSAP.com now appears to have withdrawn that article, so it's no longer possible to rely on any of the material that was in it. But neither of my key points— that SAP myCRM is hard to implement, and that AMR estimates about two-thirds of purchased seats remain unused — were revelations that haven't been substantiated elsewhere.

Then a few days later, analyst Josh Greenbaum posted a talkback comment about "all the errors in your blog" although, as I quickly pointed out, without actually citing any of the errors that he believes it contained (most of his comment fulminates about an alleged misquote in the searchSAP article that I deliberately chose to omit).

The email from SAP's PR rep went on to refute my opening assertion:

"There is NO link between the introduction of SAP CRM on-demand and implementation rate of other mySAP CRM customers. mySAP CRM and SAP CRM on-demand solution are two different products designed to meet distinctly different customer needs. There are a host of reasons as to why mySAP CRM customers purchase the suite and either wait to implement or take a phased approach to going live on the solution. It has nothing to do with the solution and its ease of use."

Notice there's no denial there that many mySAP CRM seats are lying idle — albeit for "a host of reasons." In fact, research from Gartner indicates that I may have been overgenerous to SAP when I put the figure of unused licences at two thirds. In 2004, Gartner found that "only 19 percent of mySAP CRM buyers actually use it" — which means four fifths don't. Even where it was deployed, Gartner found that most customers "have less than 50 seats in production."

But as I wrote in a Talkback response to SAP employee jnolan, other vendors are also guilty:

"What is probably unfair is to single out SAP. I wonder how many Siebel seats have never gone live? Idle seats are a scandal of ALL packaged software, not just SAP's — and CRM as a category seems to have been one of the worst offenders."

Last August, AMR Research published the results of a survey of 271 companies that 46 percent — almost half — of all purchased ERP seats remain uninstalled. From ZDNet's report at the time:

"According to AMR Analyst Jim Shepard, many customers put their internal ERP expansion plans on hold as the economy slowed, which helped contribute to the high proportion of unutilized licenses. In the worst cases, corporate downsizing might have eventually left some companies with more ERP seats than employees ... As a result of the trend in unused ERP seats, AMR said, software makers will likely shift their long-standing strategy of offering widespread volume discounts that encourage customers to buy larger numbers of licenses up front. Such a strategy change could also help quell customer complaints regarding the maintenance fees they're required to pay for unused seats."

I think if I were one of those customers, I'd be really fuming by now. Not only have I bought more seats than I can possibly use, now I'm lumbered with paying annual maintenance fees on them! This was the crux of my posting, that the business model of conventional packaged software vendors means they concentrate on selling licences no matter what, whereas on-demand vendors only get paid for live seats. As I explained when spelling out What makes on-demand better, the pay-as-you-go model creates a powerful incentive for vendors to make sure users get productive as quickly as possible with the application.  What incentive does SAP have when it's already banked its licence fees and is earning maintenance on software that's sitting on a shelf?

SAP may not realize that it's being compelled to enter the on-demand market because its conventional software is being sold where it's not needed and is too hard to implement anyway. But those are the market dynamics, and until SAP actually does face up to them, it's unlikely to produce an on-demand alternative of its own that properly addresses these issues — which is why its current offering is such a disappointment.

So this is what I wrote last Thursday morning in reply to the email from SAP's PR representative:

"I would be happy to get more information on the topics I raised in my piece. There are two specific points:

"1) Customers find it takes a long time to get mySAP CRM up and running as an on-premises solution.

"2) Many purchased enterprise software seats lie idle, but SAP CRM seems to be particularly prone to this.

"I am not clear which of these facts you are disputing.

"The conclusion I went on to draw about the link between these facts and SAP’s introduction of the on-demand offering is indeed a matter of opinion rather than fact. It is a point I make to underline one of the key benefits of the on-demand model, which is that payment is directly related to usage rather than purchased seats. I would be surprised, though, if this were not one of the factors considered by SAP when planning the introduction of its on-demand offering."

I have not yet received a reply.

PS: Jason Wood puts a measured counter-argument to my diatribe in Memo to Phil Wainewright...Shelfware isn't Scandalous. What do you think? Am I getting steamed up over nothing? Does no one care about shelfware?