Last Friday I listed the competitive strengths of different types of CRM vendors. Turns out that eWeek (www.eweek.com) had something similar on its mind. “SaaS: More, not less, channel” (eWeek, November 27, 2006), available here, argues that hosted systems will move beyond their current advantages of low cost and rapid implementation. The new positioning will draw on expert channel partners such as Value Added Resellers to help customers not simply install the software, but also train their staff and modify their business processes. Channel partners will also use the hosted systems as platforms for multiple applications and industry-specific “vertical” solutions. The eWeek article quotes, and heavily draws from, a report “SaaS 2.0: Software-as-a-Service as Next-Gen Business Platform” from Saugatuck Technology (www.saugatech.com).
Sauguatuck seems to have done its homework, interviewing 40+ executives, surveying another 155, and speaking with more than a dozen vendors and investors. Perhaps as important, what they say makes perfect sense: all those experts selling and supporting on-premise software won’t go away just because hosted systems become more popular. Nor will end-users’ need for support suddenly vanish. So it’s reasonable to expect that the experts will shift to servicing the hosted platforms as demand moves in that direction.
But I don’t find this a pleasant prospect. Even in the on-premise world, services easily account for more than half the total cost of deploying packaged software: a typical multiplier is 3 to 4 times, and 10 times is not unheard of. Very little of that work is the software installation itself—most is in customization, integration and training. If the same channel partners provide the same services for on-premise solutions, the costs can’t change by much.
Of course, companies that really need a lot of customization and integration won’t have much of a choice. But one promise of hosted systems has been that users would be able to do more for themselves. Less customization would be needed because the systems could be configured without code changes. Integration would be simplified by standardized connections using technology like Service Oriented Architectures.
You can argue that even these sorts of configuration and SOA connections are beyond the capabilities of many organizations, and thus expert assistance will still be needed. That’s probably true to some degree but looking at what amateurs can do with things like Web mash-ups, it’s reasonable to hope that the tasks can be simplified to the point where users can do much more for themselves, and experts will accomplish the remaining work in a fraction of the time they now take with traditional technologies.
Here’s the problem: if the hosted vendors depend on channel partners to sell their systems, they may not invest in features that cut into those channel partners’ income by letting end-users do more for themselves. I recognize this sounds a bit paranoid, but channel partners do play a large role in developers’ decisions. The good news, I think, is that there will always be new hosted competitors without a large channel partner base. These firms will have every reason to keep their products simple. So hopefully the market will provide a range of solutions that let some users do things for themselves while making help available to other users who need it.
As to the business process redesign component of channel partner services: to some degree, that can be built into software. To a greater degree, though, I suspect that will always be something that requires significant human expertise. But process redesign and system deployment are quite different skills, so there’s a hope that simple-to-deploy systems will allow companies to hire process redesign specialists who are not also systems integrators. Allowing more firms to offer redesign services should increase competition and ultimately serve end-users better. But this also depends on the hosted software developers keeping their products simple enough that non-technicians can deploy them.
No comments:
Post a Comment