If there's one phrase I've used at the conference yesterday and today more than my usual word -- "exciting" -- it's "the shackles of success." More specifically, I find myself using it when we talk about RedMonk's IBm dead horse: they should be doing SaaS a whole lot more, across the board.
The problem, primarily in the SMB space, with that suggestion is the large partner and ISV network that IBM's build out over the years. As with most companies that sell "stylized middle-ware" (as I'm told one IBM'er calls it), IBM sells quite a bit through other people rather than directly. This is effectively 100% true in SMB.
The problem is, SaaS offered directly from IBM could cut out most of the margin for IBM's partners, if not cut them out entirely. Pissing on your partner's and ISV's shoes, let along stealing them, is obviously a huge no-no.
The problem for IBM, though, is that other people -- like the theoretical "Google Enterprise" -- don't have partners shoes to worry about, so they can start out fresh. That is, the looming threat seems real: if end-user's understanding of how they want to use software changes from behind-the-firewall, packaged to SaaS, hosted, IBM and their partners could find themselves behind. And that'd be even worse than the afore mentioned loafer micturating.
From some conversations, it sounds like IBM wants to move closer to an SaaS model, and they're even working on which business case(s) they'll go with to keep the partners happy. They tell us it'll be great ;>
As ever, though, the question is how long do they have to wait. I'm no Kreskin, but it seems like 2007 is the year to figure it out before it gets too late.
Previous Topic
