The Royal Bank of Scotland (RBS) took over NatWest bank last week and IT costs were a major factor in the takeover battle.
The RBS infrastructure consists of four mainframes across two fibre-linked sites in Edinburgh which handle 2,500 mips using 20 terabytes of storage. Following some outage problems in l997, the bank has used consultancy Compass Reference Group for the past three years.
During that time, it has saved £12.5m each year, and reduced its running costs to two-thirds of its 1996 figure, despite the fact that the workload has nearly doubled.
Although optimising technology remains critical in achieving cost efficiency, the savings were not mainly driven by technology changes, explains Jeff Chalmers, head of central infrastructure management, IT services. The first improvement was procurement practice.
"I've had my key managers trained in negotiating skills, tackling both hardware and software acquisition," says Chalmers.
He is confident that he now has a team that secures the best deals in the UK with the likes of IBM and Hitachi, and d irect access storage device suppliers. But Chalmers believes hardware procurement is the easy bit. Even with a new software portfolio management programme specifically targeted at taking cost out and reducing the number of suppliers, "some software suppliers are extremely difficult to deal with", he says.
The key to good software portfolio management has been to reorganise lines of responsibility internally, says Chalmers. Rather than have a service manager looking after a particular line of hardware, such as Tandem, and all its associated support and software costs, RBS now considers the entire usage of a particular supplier's software across the datacentre.
Formerly, no one looked after the total spend with one supplier. "Now we can leverage a bigger pot, and we've found that very effective," says Chalmers.
He acknowledges that having to manage the datacentre's software portfolio is an overhead, but says it pays off handsomely. He told one recruit to the job that if he managed to save a single pound over his salary, he would be worth it.
"He took a half-a-million pounds out of our software costs," says Chalmers. After the initial big wins, year-on-year improvement is more difficult, "but we're still seeing success", he says.
The second major contributor to the datacentre's cost savings has been the reduction in the number of its contractors. Chalmers ack- nowledges that contractors are necessary when it comes to fine tuning the peaks, but when they end up doing virtually permanent support jobs, it's time to rethink, he says. "Some contractors had been here longer than permanent staff," he points out.
The number of contractors has been cut by 75%, by converting the individuals to permanent staff, or simply doing without. RBShas redeployed contractors that are surplus to the datacentre to business development projects, thereby achieving significant cost savings.
Another non-technology source of cost reduction has been in change and problem management. This, says Chalmers, was deemed to be less efficient in the datacentre than elsewhere in IT.
"There was not a problem in the incident management; it was getting to the root cause and identifying fixes that we seemed to be devoting more resources to," says Chalmers.
But although he is investigating the discrepancy, Chalmers is also aware that the issue raises the question of overall cost efficiency. If, as indeed happens, the datacentre achieves 99.8% efficiency on its mainframes, will making cuts in problem management endanger that?
"It can go in swings and roundabouts," warns Chal- mers. Penny pinching in one area can waste pounds in another. It's essential to understand the whole complex equation of dependency and cause and effect within the operation. Getting the balance right between maximum cost efficiency, and dangerous exposure to risk can be, says Chalmers, like "drawing a line in the sand".
He's already facing that issue over the question of investing in Parallel Sysplex, wondering whether any extra investment cost can be repaid by moving workload around the configuration dynamically, exploiting all the mips available while cutting down on downtime.
If that extra capacity should be needed on a particular occasion, one way that Chalmers covers risk is by using an IBM facility which allows users to fire up the extra Cmos engines within the four existing mainframes. Although the service costs money, "you don't pay [for the extra mippage] until you need it", he points out.
Chalmers also knows that, however ace his team's negotiating skills, there's still money to be saved by exploiting new technology. With the price of storage still dropping 25%-30% a year, "we're looking at storage area networks, and we want to eliminate manual tape mounting completely".
RBS has proved to be efficient in its streamlining of systems, but old technology can still be run highly efficiently. RBS enjoys the distinction of being the most cost-effective user of microfiche in the world, an accolade it wasn't aware of until Compass pointed it out.
"It came as a surprise," says Chalmers. "We didn't know we were so good!"
One good thing, he adds, was that many of the staff who work in the microfiche area are not the department's top earners. "It was nice to take the lower-paid staff out to dinner [to congratulate them on running] a very slick operation," he says. Microfiche may not be the most exciting part of the datacentre's operation but "it all helps the overall picture", says Chalmers.
In spite of these achievements with cost effectiveness, the perpetual pressure on Chalmers is not just to keep the datacentre running at the lowest cost and highest availability.
The third principle of operation is responsiveness. "We have to be able to change quickly and introduce new business functionality, even though change is cost," he points out.
In banking especially, change can be the order of the day. More than ever, it's essential to combine doing the day job with keeping costs down and capability high.
"If you get the bread and butter wrong, there's no point talking about e-commerce," warns Chalmers.
But for the moment, Chalmers has his mind on points south. "My simple aspiration is to work with our Nat West peers and get up to the same level on a grander scale," he says.
At least he can be grateful he doesn't have to integrate the IT systems between the two banks - a poisoned chalice if ever there was one.
Cutting costs: the RBS way