Down on the Server Farm

A growing insurance company finds that its IT resources are stretched during peak times. Can utility computing help it concentrate on its core busi...
Arthur ClennamApril 28, 2004

BM’s Universal Server Farm (USF) in Shatin, Hong Kong, looks more like a control tower poised for heavy traffic than a place designed for cultivation. TVs in its spartan control room scroll bewildering screens of data. Cavernous rows of server racks groan with metal boxes, each with blinking lights. Technicians scurry past bearing clipboards and cellphone nerd packs, and the atmosphere is quiet and intense. To Garry Willinge, director, IBM Global Services Hong Kong, the USF — one of several in China — forms Asia’s advance guard for “e-business on demand.” “We call it transformational outsourcing,” says Willinge, “and the idea is that the process will allow clients to concentrate on what they do best: their core businesses.” He adds: “We think it has tremendous potential in Asia, where young companies can buy into the model as a means to support sustainable growth.”

Utility Infielders

Is this, as Willinge would hope, Asia’s dawn of utility computing? Despite its forbidding and unromantic name, utility computing is attractive in its simplicity.

Drive Business Strategy and Growth

Drive Business Strategy and Growth

Learn how NetSuite Financial Management allows you to quickly and easily model what-if scenarios and generate reports.

Its goal is to allow users — read CFOs and CIOs of companies that want to be rid of those expensive mainframes — to plug into a remote ‘farm’ of servers, churning up just as much computing capability as they need, for just as long as they need it, and paying only for what they’ve consumed. As with traditional utilities, customers can make trade-off decisions about how much they pay versus the quality of service they get. Among those already in play is Hewlett-Packard’s Utility Data Center, a package of software and services designed to help companies easily manage and reallocate computing resources. IBM has pledged US$10 billion to this effort, which includes developing Universal Server Farm facilities like the one in Shatin.

The practice has made inroads in the US, and the bigger vendors are trying to extend it throughout Asia, where their competitors are plucky Indian companies that have redrawn the competitive map for IT outsourcing. In response, the major IT vendors have been struggling to package a whole variety of services that can be linked to the utility computing model. The idea is to link value-based consulting to their services under an umbrella server, often with a slightly confusing name. IBM has dubbed it “e-business on demand”, while Hewlett-Packard calls it “the dynamic systems initiative”. Sun Microsystems has applied the cryptic “N1” to the process.

Moving to Variable

No matter what name it goes by, the benefit to a CFO is to morph fixed into variable costs, and free a company from excessive management of IT. That’s what attracted Jonathan Bradbury, CFO of Winterthur Life Hong Kong, who has moved Winterthur Life’s computing power to IBM’s Shatin server farm. Bradbury saw the computer outsourcing arrangement in tidy alignment with Winterthur’s newfound ambitions to deploy IT to adapt quickly to the requirements of its market for life insurance sales by creating new products and reaching new leads.

Bradbury’s interest in the galvanizing effect of technology may stem from a background that encompasses both banking and running his own business. He arrived in Hong Kong 20 years ago as a Standard Chartered banker. He ran his own pharmaceutical business in the Philippines and was introduced to Winterthur during his next career as a compliance officer called in to advise Winterthur. After the assignment, Winterthur’s managers asked him to join as CFO in January of 2001.

Owned by Credit Suisse group, Winterthur Hong Kong is a branch of Winterthur Life. Its strategy is to become an entrepreneurial company that goes about the unit life business in a different way from Asian regional competitors. It has pinned its hopes on unit-linked, lead-based selling covering as many ‘channels’ — or types of products designed for specific income groups of consumers — as possible.

Supporting New Strategy

Bradbury saw Winterthur as well positioned to differentiate its business against a crowded market for competition in Hong Kong. “Businesses in insurance, including the banks,” he says, “were thinking along the same lines, going through the same cost equations, looking for assets under management, and focusing on the middle- to upper-income market.” What Bradbury wanted was the “ability to go across the whole spectrum of Hong Kong, through the various channels that Winterthur had.” But he needed both more power and less cost in IT to get this model off the ground. He says: “I thought, ‘If we can customize systems to meet with the ever-changing needs of distribution channels, we could develop a longer relationship with the customer.’” He wanted to move selling life insurance from a “push business to that of a pull business.”

Bradbury began searching for an appropriate outsourcing model that would take fixed costs and operational responsibilities of running mainframes out of Winterthur’s hands. But there was a more important consideration. Winterthur purchased a Siebel CRM platform in 2002 to support the new strategy, and it immediately began stretching the company’s computer capacity and hardware management skills. Further, Winterthur suffered from the need to maintain excess IT resources due to CPU activity during peak periods. The activity typically spiked twice a month, but remained low for the remainder of the month.

“As you grow and you sell more policies, CPU activity and storage requirements increase,” says Bradbury. “When you sell more policies you have more premiums to collect and entries to post. What you have are two big spikes in a month, when the CPU needs get bigger and bigger in order to process the increasing number of transactions within our fixed offline back up period.” These spikes could last two or three days. The demands on the CPU were getting so large that they threatened to encroach on online daily activity, an intolerable situation, Bradbury says.

In the end, the 23 additional servers needed to run the Siebel CRM coupled with increasing CPU activity spikes on Winterthur’s IBM AS/400 mainframe called for extra data center capacity which would add to fixed costs, including new space, added staff, and improved temperature and humidity controls.

Bradbury shopped around, but it was “IBM that came to us,” he recalls, “and said, ‘Why don’t you look after your business and we’ll buy our boxes back?’” Bradbury liked the terms. Under the contract that he struck with Willinge, IBM provides all data services from the Shatin center. The contract allows Winterthur to be charged on the level of computer use, and it doesn’t pay for idle capacity. In all, Winterthur Life Hong Kong has outsourced the operation of nearly 100 systems it once owned to run its business, encompassing the IBM AS/400 servers running the life insurance policy management system; storage area network systems; plus Windows 2000 and NT servers primarily supporting a Siebel CRM system, along with sundry networking systems. IBM also sweetened the deal by buying back Winterthur’s AS/400s and incorporating them into the server farm.

Bradbury says that he set out to reduce IT costs by 20 percent, but reports that overall savings are much more. IBM has advantages, too. The Shatin USF’s customers include PCCW, the Hong Kong phone company, and JPMorgan. The greater the number of large clients, the easier it is for IBM to shift capacity from one client to another. It also garners economies of scale of all of its costs, from maintenance to the ability to consolidate all backup and disaster recovery systems.

Has it come off without a glitch? Bradbury thinks so. And Willinge says that service contracts these days have performance demands written into them that guarantee a jealously negotiated level of service.

Dynamic Utility

Still, the pundits who are watching life on the Universal Server Farm can add a dash of skepticism. “Most vendors are underplaying the complexity of the changes companies have to go through to do this kind of dynamic utility computing,” says Mary Johnston Turner, vice president and director of the large-enterprise practice at Summit Strategies in Boston. And those changes aren’t just technological, she says. “To realize all these beautiful benefits — saving money, improving productivity, making IT responsive to the needs of the business — you really have to adopt a policy-based automated services management approach” as well. That will mean, among other things, striking a balance between selective outsourcing and internally run systems.

Despite those challenges, analysts predict that, by any name, utility computing makes so much sense that it will become an everyday business reality — eventually. “There are a lot of moving pieces to it,” says Turner. “I really do believe we will get there, but I believe for most enterprises, it’s at least a five-year journey.” In fact, Winterthur’s journey was nearly five years, from the time it took for Bradbury to draw up a strategy related to the outsourcing, sell it internally to the company, and finally, make the move to outsource his entire IT infrastructure.

But five years is hardly a long time in the life insurance business. Bradbury sees Winterthur now positioned to do primarily what he thinks it should be doing: selling policies to a vast, untapped market. “The advantage is that this frees up our IT budget for development work,” he says. “It gives us better control of the life business.”

Arthur Clennam is a Hong Kong-based journalist specializing in back-office subjects. Portions of this story, written and reported by Ann Stuart, appeared in the Winter 2003 edition of CFO IT magazine, published in the US.