For Utility Computing Opportunities Think Outside the Enterprise

If nothing else came out of the dotcom boom and bust we learned that connecting to customers and suppliers is not easily done with applications developed to be run inside the firewall. Many of the early e-marketplaces or trading exchanges were born out of the realization that inter-company processes can be more efficiently handled outside the firewall. Indeed, some of the services we would call utility computing today have their roots in the e-marketplace world.

Not only is it logical to use services sitting outside the firewall for connecting to trading partners, it is cheaper. Imagine a relatively small trading community with a dozen partners. If each partner connected to every other partner a total of 66 connections would need to be made. Alternatively, each can connect once to a third party service provider which would be a total of 12 connections. The connection cost for each trading partner is significantly less in the utility computing model, not including the ongoing maintenance costs.

Examples of logical services for the utility model include hosted procurement ( IBM , IQNavigator ), hosted integration ( Interlace/webMethods ), and hosted sales tools ( salesforce.com ). Industry marketplaces such as Covisint , Elemica , and Trade-Ranger also provide of mix of different hosted services. However, given their ownership structure, users should look carefully at their underlying cost structure before committing to them. In their favor they are well represented in the industries they provide services to.

Technology Decisions Are Dead, Long Live Technology Decisions

Theoretically the utility computing model frees users, once and for all, from the tyranny of keeping up with the ever changing technology landscape. Most users would prefer the vendor simply provide the functionality required when and where it is required. They would rather not dig into the inner workings of the underlying applications. After all, to make an analogy with an established utility, most of us don’t know, or care, how our electricity gets delivered to us – as long as it gets delivered reliably (not as certain as it seemed before the recent blackout in the eastern United States).

The reality today – with the utility computing market being so immature – is it’s still a good idea to look under the hood. Only an application built from the ground up as an Internet-based application will be able to deliver on the promise of cheap, reliable IT functionality delivered over the web. To a certain extent the pricing model will tell you what’s under the hood. If the service provider is willing to deliver its services on an attractive monthly subscription or transaction fee basis (as opposed to a long term contract) then there is a good chance it has built an infrastructure that can do so cost effectively. But beware of outsourcers, ASPs, and former dotcoms in utility computing clothing. They might be willing to sell their services at a loss with the expectation they can raise rates later. Looking closely at the underlying infrastructure will tell you how sustainable the pricing model is.

.

Making a Big Splash…Without Rocking the Boat

OK, so you know you want to try the utility computing model but you are still reluctant to make a big bet on something so new. The risks taken and the limited gains received by those who took the plunge into the outsourcing and ASP markets are fresh on your mind. The good news is the utility computing model is highly scalable. You can start small and grow as the value becomes apparent (but see previous section on technology evaluation).

Where you start will depend on your business, where the biggest need is, or where you can have the biggest impact on the company. If the emphasis is on growth and you have a distributed sales force, providing sales force tools is a cost-effective way to improve sales productivity. If the focus is on efficiency and order processing (either procurement or sales) is still primarily a manual process, providing procurement or connectivity services is where you’ll want to start. Bottom line: Picking a high impact utility computing project is not different from picking a high impact IT project. It needs to line up with your business opportunities and challenges.

Like any new way of doing business, taking advantage of the utility computing model is best done one business process at a time. It’s an immature market. However, unlike previous attempts to change the way we purchase and use information technology, utility computing can be approached incrementally. It’s not risk or cost free but, as you will find when you start to explore your options, it’s your best bet for making a big splash…without rocking the boat.

Leif Eriksen is founder and principal at Industry Insights, an independent consultancy. He has over 20 years experience in business strategy and technology and can be contacted on leif@industryinsights.net.