My colleague Michael Hickins and I have disagreed at times over whether companies like Google or Apple. My skepticism has never been one of simply product function or cost, but rather an issue of what it takes to provide what an enterprise IT customer looks for. Bernard Lunn at ReadWriteWeb has an interesting discussion of this in the context of enterprise customers often demanding on-premise deployment, rather than SaaS delivery. And in his reporting is a nugget that suggests an entirely new business model that I'm calling peer-to-peer enterprise computing.
According to Lunn, have the vendors he spoke to at the Enterprise 2.0 Conference delivered services on-site, versus in an SaaS or hosted model. Why? Because that's what the enterprise customers demanded. They had a few reasons. (Though how many do you need to hear beyond, "This way or no pay?") However, one of the most interesting was a matter of resource utilization:
But the figure that really tells the story is 6%. That is the percentage of server utilization in enterprise data centers, according to McKinsey. That is a lot of wasted cycles. It would be much better to use them up with new applications, and to bring in virtualization technology to use them more efficiently.Why rent more cycles from a SaaS vendor when you are swimming in excess capacity? That is such an important and sharp observation as to be beyond over emphasis. The companies have made investments and want to get as much from them as possible. I'm guessing that they negotiate a substantial price reduction as they're bearing the hosting cost.
But it got me thinking. You see some people setting up solar or wind power generation on their property and then feeding back electric to the grid, reducing their power costs in the process as they make that meter spin backwards. Why not enable the same thing for enterprise computing? Essentially, this would be treating enterprise computing like peer-to-peer grid computing for projects like the distributed grid computing project, SETI@home. The difference is that the enterprises would not be doing this for free. Instead, you'd have an international computing grid virtually deployed over the Internet.
Companies that needed a hosted app would make arrangements through their service provider. There would be two separate line items: the application license and the hosting/communications. Enterprises could either pay for hosting or make some of their excess capacity virtually available. Yes, there would be a number of thorny issues to solve: billing and credit, a central directory of availability, security, and the like. But if there really is that much excess capacity in large corporations, you'd think they could wall off parts, maybe using some form of network segmentation, to have it available for mid-sized or small companies that could benefit most from an incremental approach to adding capacity, or to service providers who needed machines to run their software.
To put it differently, corporations could create a giant communal cloud that vendors could make available. This would reduce the massive capital investment that would otherwise be necessary while offsetting the operating budgets of IT departments. And even with some of the potential difficulties and concerns, there is something emotionally transformative to a CIO, CEO, or CFO about lowering operating costs.
Image via stock.xchng user juliaf , site standard license.