I've written extensively on the disruption that cloud computing presents to IT organizations; I've focused much less on the disruption it poses for the entire IT supply chain--from vendors through to system integrators and on to outsourcing companies. Make no mistake, though, cloud computing will disrupt every part of IT. No sector of it will come through unscathed.
A couple of anecdotes that were shared with me at a conference this week brought this home.
The first anecdote outlined the challenge one large vendor was experiencing with its channel partners. Today, these partners help the vendor's customers implement and integrate on-premise hardware and software. A typical project lasts months (if not years!) and requires a small army of technical people with different skills: project managers, data analysts, interface programmers, finance and accounting personnel, etc., etc. In a cloud world, these projects shrink to much shorter timeframes and require many fewer people, i.e., many fewer billable hours. So, from the channel partner's perspective, cloud computing is much less palatable.
[For timely cloud computing news and expert analysis, see CIO.com's Cloud Computing Drilldown section. ]
The second anecdote came from an end user which had signed a large outsourcing deal. The goal of the offloading was to save money, which it accomplished. However, now it takes six months (that was the figure quoted) to get new resources provisioned. The speaker displayed visible anguish at the position IT was in: a resource provisioning process that take forever and end users with increasing demands for IT resources and an increasing impatience fostered by what they've heard about the "available in minutes" characteristic of cloud computing. Talk about a rock and a hard place!
With regard to the first anecdote, it's clear that the asset-lite, easy provisioning, transparent economics of cloud computing require much less of certain portions of typical on-premise projects. However, it's not clear that the "tip of the spear" part of on-premise projects--integration with existing systems, end user interfaces, etc., disappear at all.
It may be that if a SaaS solution is being used formerly necessary parts of projects that were part of the value of the project will not be needed--for example, traditional CRM implementations like Siebel had endless database and schema work, all aimed at creating an environment perfectly tuned to the user environment. With Salesforce, the service comes with a standard schema that customers are expected to live with--and guess what--they do.
Nevertheless, the valuable parts of system projects, even if the overall project is less costly, remain--and perhaps deliver even more value when the overall project is freed from expensive, low-value labor.
Case of Cloud and the Clipping Service
Let me offer an example. In the past, most company's marketing departments subscribed to a clipping service--a service that subscribed to newspapers and magazine and would clip stories that mentioned the company. Every week or so, an envelope would arrive from the clipping service with all the stories they clipped neatly pasted to sheets of paper. Naturally, this was expensive because the people cutting out the articles had to be paid for. And, it was difficult to justify spending more on different clips, about competitors, say, because of the expense. Moreover, it was time-consuming and a hassle to change what you clipped for (we term this process friction), so it wasn't done very often.