The end of 'one throat to choke'?

Are enterprises willing to rely on multiple vendors to deliver their most essential apps?

OK, first let's dispel two myths foisted on us by big-name software industry personalities. CEO Marc Benioff, as of Sept. 12, would have us believe that his company will become a major platform for enterprise application integration. This is not going to happen.

Next, while Oracle is building an impressive software portfolio, neither Oracle nor its CEO Larry Ellison has enough time or money to buy every single enterprise software vendor.

That said, each company is responding in its own way to the desire for more business process connectivity to the enterprise suite, says Josh Greenbaum, senior analyst at Enterprise Applications Consulting. But how do they get there? faces the same problem that Siebel faced. If the customer record, as defined in CRM solutions, is the nexus of a tremendous amount of strategic business processes, how do you connect that record with the rest of the suite?'s solution is AppExchange, a platform that allows its customers to incorporate third-party Web services and Web-services-based applications into the architecture. While it sounds like Benioff is on the cutting edge with an SOA-like solution, I believe AppExchange is counterintuitive, especially when you consider everything else happening in the market.

Instead of one throat to choke, one stack to manage, Benioff would have us believe that large companies want to go out and link dozens of smaller applications together, some of them untried and untested. That's what componentization is all about, I suppose. But I tell you, this idea runs completely counter to the real world, where companies are trying mightily to reduce the number of applications they need to manage.

When I talked with Greenbaum about this, however, he called me naïve. The idea of one throat to choke is overhyped, he says. While he agrees that no major enterprise is going to put its apps on a "dinky little infrastructure" -- AppExchange, that is -- he says no one vendor meets all of the requirements. If Greenbaum is right, then we end up with a lot of battlegrounds. And that is exactly what Evan Quinn, vice president of applications research at IDC, believes will happen.

SAP is using NetWeaver as its infrastructure parlay against all of its classic applications competitors, Quinn says. Oracle, on the other hand, is acquiring a big market applications footprint and saying to its competitors -- mainly SAP and Microsoft -- "let's see if you can follow." And don't discount Microsoft, Quinn adds. Its recently announced Dynamics initiative promises to roll all of its enterprise software into a single code base by 2008.

So we are left with SAP NetWeaver as a business process platform; Oracle's boast that it can offer everything from accounting to fulfillment, logistics, and more;'s Web services platform for third-party components; Microsoft Dynamics; and possibly IBM's Workplace middleware, which offers everything but the applications. Quinn says it will take many years for the market to figure out who's right and who's wrong -- not much solace there as you finalize your IT budget for 2006.

My guess? Pure-play on demand solutions will gain only limited acceptance in the enterprise. SOA will become a popular internal solution but, along with the so-called extended enterprise that drives it, will never quite get off the ground. After that, my crystal ball gets awfully hazy.