The Bloor Perspective: Procurement outsourcing, ROI and the PS2 grid

There was widespread anticipation earlier this month that Deutsche Bank is planning to outsource its procurement operations worldwide. It is estimated the move will save E700m, or about 10 per cent of its annual spend of E7bn on goods and services.

Deutsche Bank has already outsourced its IT operations and this is a further step in its initiatives to focus on core business, reduce costs and improve profitability. It is speculated the outsourcers will be Accenture, EDS and IBM.

Substantial elements of procurement are devoted to IT and telephony products and services. It is assumed the recipients of the procurement outsourcing deals are not those who have benefited from the outsourcing of the IT operations.

There must be an inherent issue in the same organisation running and procuring technology products and services. Doubtless, the functions are handled by distinct and separate entities or organisational units within the outsourcer enterprise but the perception must prevail that they will favour their own and use their combined power to obtain the best deal.

The combined power of the operator and procurer may certainly extract the keenest pricing and service. Would, for example, IBM as a procurer have any restrictions placed on its selection of IBM services and products? Should procurement and IT operations in effect be seen as different facets of the same business. In the context of IT, the procurer may be seen as the purchasing arm of the IT operator.

It will be interesting to learn from Deutsche Bank - if the outsourcing procurement deal is indeed announced - how IT procurement and IT operations will be positioned and reconciled with one another.

*ROI is king*

With the slow economy, return on investment (ROI) has again become the latest industry buzzword. Touted by vendors keen for coverage it is almost as though they have suddenly found the Holy Grail. And yet the concept, a full explanation of which is best left to the accountants, has been around for years. So if we've ignored it in the past, why do we care now?

Most projects start with a business case and high expectations. Someone, somewhere has identified a need. Everyone gets involved and it's all quite exciting. Expectations are set while everyone is on a high, plans are laid out, roles and responsibilities set and off we go.

Things start happening quickly. Normally change is the first thing to happen. As you move forward, you find there are more problems than originally thought and risks and issues come at you from every direction. You quickly realise that this thing is bigger than anticipated and you need help.

Sometimes you get internal help and sometimes you need to pay for outside assistance. While this is going on other things change - not just someone moving the goal posts but throwing more balls on to the pitch.

After the initial interest and involvement the high level stakeholders decide that you are best left to your own devices and should only call them in when absolutely necessary. No doubt when you do they will be too immersed in the next new project to give you the time you need.

As time moves on, the changes increase and you have to counter. Instead of things being added to the project you start to take them away in an effort to meet an already impossible deadline. Once people find out that you're not addressing all of their needs as promised, they too find better things to do with their time.

At the project inception, you have visions of testing and documenting the whole system. As time goes by reality dawns - design documentation is not kept up to date, and any other documentation just doesn't get a look in. Testing is compromised by time restrictions, an ever-changing scope and loss of interest from key parties.

Ultimately, something usually gets delivered. It is often not what you expected but you've gone live. Before you've popped the champagne corks you have to immediately embark on a maintenance programme that fixes the faults found as a result of inadequate testing or to add the functionality that you chopped out of the scope to deliver on time.

At no time have you considered the original business case. You've probably lost track of all the costs and changes and as far as measuring the ROI - who cares, the systems in and that is the end of that.

*Grid games? Getaway*

The concept of grid computing has received much attention over recent months, as vendors such as IBM and Sun have attempted to move this new technology out of the academic world into everyday commercial use. This month came news of a development that saw IBM and Butterfly.net, one of the pioneers of grid gaming systems, teaming up to supply technology and networking services to Sony Computer Entertainment.

Under the agreement, Butterfly.net and IBM have created the Butterfly Grid, a network gaming environment for Sony's PlayStation 2. The system has been designed to allow online video game providers to deliver sophisticated games to millions of concurrent users with very high reliability.

Until recently video games have usually segmented players onto separate servers thereby limiting the number of gamers who could interact with each other. This architecture also raises considerable support obstacles as suppliers seek to ensure that the games are highly available.

It is not an uncommon experience for online gamers to find that when one server shuts down or is overloaded, game-play comes to a halt. Using Butterfly's grid technology server, interaction is completely transparent and seamless to the user creating a gaming infrastructure where servers can be added, or replaced, without interrupting game-play.

The Butterfly Grid is hosted and managed by IBM and operates using dual Intel Xeon blade servers running Linux. IBM's WebSphere and DB2 are combined with Butterfly.net's systems into an integrated platform for online game development, deployment and ongoing operations. The grid has been designed to support over one million simultaneous players from each facility in a 24/7 environment with automatic fail over.

This development highlights the commercial exploitation of grid-type infrastructures is now becoming a reality. Indeed, as well as extending the scope and raising the quality of service already offered in the internet gaming world, it is clear that Butterfly.net, IBM and Sony clearly see potential in grid technologies to support entirely new revenue streams.

If successful, this move by Sony will allow large numbers of people to make enjoyable use of grid computing without even being aware of it. Just the way good IT should operate.

Bloor Research is a leading independent analyst organisation in Europe. You can find out more at www.bloor-research.com or by emailing mail@bloor-research.com.