In service

Service management has become a given for IT executives. For the past few years, almost without exception, IT directors, CIOs and other technology decision-makers have voiced their commitment to orchestrating the underlying technology infrastructure so that it supports and aligns closely with key business goals.

The problem is that the reality of IT service management does not always measure up to the aspiration.

For one thing, during 2002, organisations that are already charged with delivering many of the core pieces of IT as guaranteed services have come under intense pressure to make hefty cuts to the cost of doing so.

It’s a simple formula, says Tim Greiser, an analyst at technology strategy adviser IDC. “IT is being pressurised to do more with less,” he explains. The question is: Are there products out there that can help them achieve both? With some reservations, Greiser believes so. “Increased automation can help offer operational efficiency and simplification. With ever-more systems under management, operational efficiencies can only be achieved by increasing the number of elements that can be managed and controlled by an individual IT administrator or domain expert,” he says.


In practice: Nationwide

As the UK’s largest building society, the Nationwide has a formidable infrastructure. It processes, on average, 1.5 million transactions every working day, but at peak times transaction rates can soar to three million a day. And it is willing to show its ability to run that infrastructure in a very public way. Through high-level service management software, anyone logging on to the Nationwide web site can get a real-time status report on the availability of its Internet banking services.

That is not easy given what sits underneath. A highly heterogeneous environment that includes mainframes and open systems from IBM, ICL, Unisys, HP Tandem and Compaq, supports customer services to manage loan applications and other services, as well as running all the applications needed for Internet and mobile phone banking. It is also geographically dispersed, with multiple servers distributed across the UK in the Nationwide’s 681 retail branches.

Given that, the organisation realised that with the increasing take-up of Internet and mobile phone banking among its customers, it would need to upgrade its network backbone. It selected a Cisco-based network and decided its network management system needed examining at the same time so it could provide its infrastructure operations teams a consolidated, executive overview that would allow them to quickly identify and prioritise potential service or security-related faults.

The company chose Netcool from network management software company Micromuse to provide that role of monitoring and managing its complete infrastructure from a single point. The choice was influenced by the platform independence of the product – it was capable of taking status information from any of the Nationwide’s systems or networking platforms, as well as its existing systems management platform, BMC Software’s Patrol suite. Netcool was chosen over other products, including those from BMC, because of its security and WAP features. While the security features enable network managers to use a single dashboard to isolate any threats and take action, the WAP capabilities provide real-time monitoring of the mobile banking delivery infrastructure, enabling staff to determine the availability, response times and usability of the WAP-based services and applications.

The Nationwide wanted to build resilience into the network to ensure there was no single point of failure, so that if an error occurred anywhere on the network, either through a fault or a security attack, then this would be contained and prioritised according to key business services. The upshot is a very public display of service levels. A risk, maybe, but a risk that makes any casual browser of a service management monitor.


Historically, many of the traditional tools here – for tracking the status of systems, networks, applications, as well as the end-users experience – are effective for managing separate devices and discrete activities. But while they provide a wealth of data, these tools often fail to relate that to the business processes served by the systems they monitor.

The software, however, is moving in three distinct directions that are aiding that process of achieving both increased efficiency and reduced costs. Core products from companies such as Hewlett-Packard, IBM, Managed Objects and Micromuse now provide ‘dashboards’ that consolidate performance data from the systems and software that lie behind a service and give a high-level view of the health of that service.

“The new breed of infrastructure management systems allows the IT department to present its operations as a service and to demonstrate in real terms the value it contributes to the business,” says John Holden, an analyst with the Butler Group. Organisations want to go beyond the capability of just managing the individual nodes, he says. They want to map business processes and services to the alert. So if a server is down, they know what service that failure will affect and what function the server has within that service.

The overriding requirement is to add business context to that management data, and in particular, for a platform that pulls together the data generated at device and systems levels, defines those in terms of key business performance indicators, and relates that directly to the priorities and concerns of the business – whether that is managing an application for peak performance or ensuring near 100% uptime for a specific service.

On the defensive

The systems and networks service management tools are trying to measure are certainly more complex, as are the aspects that require managing. Many of the systems that have been implemented in recent years were put in place to provide tactical solutions to urgent organisational problems, says Paul Mason at IDC. “But these systems, networks and applications were typically not deployed by following any master plan. Rarely has the problem of managing them after deployment been seen as anything but an unfortunate overhead expense. When the budgets for tools required to solve these problems and the staff salaries needed to operate them were presented to corporate and line-of-business management, IT quickly found itself on the defensive by being viewed as an excessive overhead,” says Mason.

To get a holistic picture of the quality of the service, management platforms need to span the management of the helpdesk, security management, service level monitoring, capacity management, change management, availability management, asset and configuration management, and other areas.

However, not everyone takes that end-to-end approach. Indeed, a recent survey of 200 IT managers, conducted by Benchmark Research, observed that “most companies fail to go beyond the basic first steps of IT service management.” When it came to service support, 60% declared they had an IT service desk, and just over half said they managed IT incidents formally. But only between a fifth and a third tackled such areas as change management, release management or configuration management.

The same pattern was evident with service delivery, with only 37% systematically managing their service levels, and one-fifth practising capacity management or availability management. Moreover, the survey found that only 45% of organisations formally measured the performance of the IT services delivered internally, even though two-thirds had service level agreements in place. “Although steps are being taken to improve service support,” report Benchmark’s researchers, “the procedures that are leveraged are reactive as opposed to proactive.”

The commitment seems to be in direct proportion to the size of the organisation’s management task. “Certainly within large organisations people are aligning IT assets with business objectives,” says Paul Johns, vice president of global marketing strategies at Micromuse.

When a ‘best practices’ approach is taken to service support and delivery, via policies, benchmarks and standards, such as BS15000 and the UK-government backed Information Technology Infrastructure Library (ITIL), though, there are clear benefits in terms of efficiency and costs, according to companies that have gone down that route.

Justify your existence

Organisations that have aligned their infrastructure with the business do report some significant returns on their investment. At the Halifax financial group (now part of Halifax Bank of Scotland), IT problems fell by 39% at a time when the organisation was able to increase IT changes by 125%. It also reduced the ‘mean time’ to identifying IT failures by 20%, the time to problem resolution by 27%, and the time to recover by 52%. The fundamental result: “The implementation of service management processes allowed us to measure our performance and justify our existence,” says one manager who worked on the multi-year Halifax project.

In practice: Barclays Global Investors

Barclays Global Investors (BGI), the world’s largest institutional asset management group, has a network that is dauntingly critical to its operations. The 2,000 clients who trust $530 billion in investment funds to BGI expect to be able to place orders online and manage their accounts at any time. But, like other organisations, BGI has been under pressure to take cost out of the running of that network and increase its efficiency. By introducing a single network management infrastructure, centred around the OpenView systems and network management suite from Hewlett-Packard, the company was able to cut its network management staffing levels by 12% to 30 people, even as the business was growing by around 10% a year.

In terms of network troubleshooting, less than one full-time employee now performs tasks that previously took the equivalent of 3.5 employees, says BGI. The job of supporting and tuning server operating systems, which used to require the full-time attention of nine staff, is now carried out by six employees. And the running of batch jobs, capacity planning and disaster planning and recovery is about 20% more efficient, says the company.

Perhaps most significant of all, OpenView is credited with enabling the company to reduce downtime by around 40%. The benefit of that is stunning: an average 5% increase in productivity across the company’s 2,000 employees, according to John Nichols, BGI’s global data centre manager. “The increase in availability alone has repaid BGI’s investment,” he says.

Avoiding downtime is critical to BGI, not only because of the productivity effects but also because of the revenue impact. “At the wrong time, a single hour of downtime could cost us millions,” says Nichols.

He cites the case of one critical application where the consolidation of management information has had a major impact. BGI operates multiple stock transaction databases across the company, as well as at its external brokers, which track the status of market trades. Reducing response times for that application and maintaining it in a healthy state has obvious benefits for clients. But to get the status of applications, explains Nichols, the company needs to correlate information from multiple systems. Clearly knowing the health of an application is much more valuable than just knowing the network is up, he says.



IDC’s Greiser suggests that, aside from refining processes and implementing standards, the technologies for service management are maturing to deliver even greater savings. As organisations evolve from manual procedures to automated procedures and on to ‘self-managing’ or ‘self-healing’ approaches, systems management will handle an increasing volume of the more routine and mundane events, providing rapid response to ordinary conditions and leaving IT staff available to focus on complex or critical tasks, he says. “Ultimately, self-managing systems will aid in improving availability and performance through rapid response and by reducing the number of exceptional conditions that must be handled by skilled IT staff.”

The pressures on organisations to automate routine management tasks and cut costs in the infrastructure are being reflected in an increase in spending on service management software – something of an anomaly in a period when most other segments of the software industry are in deep recession. Hewlett-Packard’s head of software for Europe, Cesare Capobianco, reports that revenues from the company’s OpenView systems and network management software business are growing at a double-digit percentage. And business service management software specialist, Managed Objects, which counts Bank of America and AT&T among its customers, has been seeing growth of 64% year-on-year.

And the reason such customers are spending is evident in some of the reports of relatively rapid and sustainable return on investment.

IDC recently carried out a return on investment audit on 14 major organisations following their implementation of service management packages, mostly HP OpenView. These included Texan gas utility company Atmos Energy, Wingcast, the in-vehicle wireless mobility joint venture between Ford and Qualcomm, and asset manager Barclays Global Investors (see box, In practice).

Its findings were truly stunning. “The companies were able to reduce downtime by an average of 79% and realise an average ROI of 1,296%. The savings achieved allowed the companies to quickly pay back the investment required to purchase and install the systems management tools.”

Specifically, the average annual savings from improved system management productivity were $7,088 per 100 users; the average annual savings, over a three-year period, from management efficiency improvements were $13,331 per 100 users; due to better system and network availability user productivity increased significantly, producing a saving of $34,500 per 100 users; and lastly, the respondents estimated that the savings from reduced downtime averaged $27,400 per 100 users. The total ROI over three years amounted to $223,585 per 100 users.

That kind of success has a direct payback for IT executives. “Raising the level of IT management from infrastructure and operational management to service management has the added advantage of raising the credibility of IT managers with the organisation,” concludes Paul Mason at IDC. “Integrating enterprise management tools can be expensive to purchase and implement, but they can yield extremely rapid returns in savings in IT management efficiency and productivity, as well as reduced system and application unavailability.” “Thus, the investments made by IT can readily be seen as creating real value for the business, making IT a hero instead of a burden,” says Mason.

Avatar photo

Ben Rossi

Ben was Vitesse Media's editorial director, leading content creation and editorial strategy across all Vitesse products, including its market-leading B2B and consumer magazines, websites, research and...

Related Topics