It seems like everybody that has written about this report always starts with, “Glad that McKinsey did this report, BUT…” Given that enough people (i.e the industry at large) has said their thanks, lets focus on the issues:
- Most customers adopting cloud computing “hide” behind cost saving to “bypass” operations. Why? Because IT operations as are very slow in responding to business needs.
- NO mid to large enterprise is even remotely thinking about moving their datacenter to the cloud. Most projects are either new, that can be done entirely in the cloud, or hybrid, extensions of existing applications that need additional features. New cloud applications are mostly driven by SaaS and customer/partner centric applications. Hybrid applications are mostly driven by the need for elasticity. Please remember this is about mid/large enterprises.
- The 15% cost savings on labor cost is wrong. The cost of running a cloud over time is much lower than running individual datacenters. Additionally these calculations do not factor in the costs from time to market.
Cloud computing is still maturing. This is EXACTLY how SaaS started out… SLAs are not good enough, it is for small companies, etc. Now we have large companies like BoA, Japan Post doing multiple thousand seats on Salesforce.com. Yes SLA needs to be higher, yes the pricing has to be aligned for continuous/long running applications, yes we need to figure out security/compliance issues (plug for Sonoa’s Secure Cloud Gateway 
), but that will come. The savings or revenue just on the time to market is big enough to justify the movement.
