Taylor hits the nail on the head:
Device management is but one example. Fixed-mobile convergence. Mobile data network roaming. "Personal" services on "business" devices and vice versa. Each and every time, the carrier service is designed with virtually no insight or knowledge of the way IT departments intend to do the same things.
There are a few examples, like the enterprise-mobile integration of PBX services (Avaya, Tango Networks, and Cisco/Nokia come to mind). But for an example of the lack of enterprise knowledge within the carrier supply chain, there is the OMA DM working group.
A few months ago, a small group within OMA proposed an extremely early-stage idea of putting security credentials onto a management object to be supported by OMA DM. This was a proposal to assemble a proposal -- as in the very beginning of the process for the Secure Element Management Object (SEMO). Within weeks of the initial idea, several very large mobile operators (Cingular and Vodafone to name a few) shot down the proposal, saying that (1) other standards organizations like 3GPP/3GPP2 and ETSI have already created standards to address the intention of the SEMO, and (2) SEMO would actually decrease security by opening the door on what is otherwise a secure exchange between the mobile operator and its customer. The SEMO objection was rapid and valid, because everyone involved in the OMA readily understood the relevant technologies and approaches.
The real issue here is that carriers aren't focused on business customers. It's not their core competency. They're good at selling service and running their networks, not at designing and building complex solutions for business IT networks.
That's why, as Taylor points out, issues such as device management may be new to carriers, but they're old hat to IT departments.
Carriers, start listening to enterprise IT managers. They know what they want and they will tell you, if you take the time to ask them.