Ref: z2615194m
Showing exactly how much money that is lost by a company when its network either 'goes down' or is simply 'slow' is an elusive science. The only certainty is that it can cost a great deal of money. For example, companies that rate cost of downtime by lost business have given estimates of $15 000 to $100 000 per hour - even up to $100 000 per minute for realtime transaction systems. This article aims to illustrate an ROI (return on investment) model designed to equip network managers with objective data that is particular to their own operations. They can then make appropriate decisions regarding the impact of their investment in a centralised IP-network analysis system.
Factors that should be taken into consideration when calculating ROI
It is generally known that a state-of-the-art centralised IP-network analysis system can help large, multinational enterprises to decrease costs through improved labour management, such as reduced 'windshield time' for support engineers. At the same time, a centralised IP-network analysis system can help increase revenues through improved quality of service (QoS) and satisfaction, for both users and customers of the enterprise network.
Investing in a modern centralised IP-network analysis system is a logical and financially sound decision for a large enterprise, when compared to the purchasing of traditional, dispatched network analysis instruments. There is, however, a need to quantify the resulting benefits. The goal is then to use the results from an ROI calculation to determine the payback time derived from the deployment of the new system. Finance directors and CFOs want to know exactly how a centralised IP-network analysis system will impact on all areas of their business operations - What will be the CAPEX (capital expenditure)? How much will it decrease their OPEX (operating expenses)? How much will the revenue increase? When will they see a return?
Centralised IP-network analysis systems affect revenue by having a direct impact on end-user productivity, through better service quality, shorter activation time of new applications and services, and more rapid and accurate monitoring of mission-critical communication links and applications. The result of increased customer satisfaction and faster service introduction is, without doubt, revenue growth.
Centralised IP-network analysis systems can also decrease costs by reducing network downtime, the number of engineer-visits to sites ('windshield time'), and by providing IT teams with tools that make up-to-date and accurate information available to them online. This information lets them optimise usage of the available network resources, and communicate more effectively with users, managers and peers.
Specifically, network engineers are enabled to perform timely and correct provisioning of what users ordered, automatically aggregate information required to manage SLAs, and perform service testing in minutes instead of days. Centralised IP-network analysis systems also reduce the cost of staff training, materials and supplies associated with staff transfers. Consequently, any ROI model should be capable of considering all of these cost impacts when quantifying the rate of return.
Since centralised IP-network analysis systems impact several aspects of the IT operations of large enterprises, a successful ROI model should be capable of measuring the impact on the enterprise's entire financial operation. From this perspective we will now look at revenue advantages and improved cash positions.
Revenue advantages
We now examine two of the revenue advantages associated with the implementation of a centralised IP-network analysis system:
* How to build a competitive advantage on innovation - accelerate data service introduction and expedite problem resolution for new services.
* How to build a competitive advantage in the delivery of quality of service (QoS).
Innovation: accelerate data service introduction and expedite problem resolution for new services
With the onset of competition and the need for increased efficiency, the focus of network managers has moved away from the network toward end-user and higher-layer service management. Offering the user population specialised services that meet their unique application needs is a key success factor for any enterprise in a highly competitive marketplace.
A centralised IP-network analysis system enables network managers and service planners to quickly provision modern, distributed application packages or network links, and efficiently introduce distributed services to the user population.
By using distributed data acquisition devices, centralised IP-network analysis systems can collect information from the network in realtime, based on a configurable collection policy. The data is then processed and stored in a centralised repository for various analysis purposes. This means that centralised IP-network analysis systems provide information that may be used at the service and network management level to deliver higher-quality services faster, or to provide distributed services, anticipating end-user needs.
Data collected from several network operators shows that resolution of a complex network problem requires an average of 380 minutes in a large enterprise. By using a modern, centralised IP-network analysis system this time falls down to an average of 90 minutes, structured in the following way:
By assuming an average fully loaded cost of $50/hour for a network engineer, the savings can then be estimated as being in the region of $250 per trouble ticket.
For the purposes of ROI calculation, experience shows that centralised IP-network analysis systems can increase the rate of new service introduction for a typical enterprise by at least 30%. The estimated benefit of accelerated deployment of new services and applications is innovation at all levels in the enterprise and, ultimately, competitive strengths.
How quality of service translates into increased productivity:
Delight end-users
For large enterprises, quality of service means less network 'slow time', no network 'down time', faster response to end-user demands and proactive management of application and service requirements.
A better quality of service, made possible by an effective centralised IP-network analysis system, translates directly into customer satisfaction, which ultimately affects revenue due to improved productivity. It also empowers network managers to increase the acceptance level of innovative services, such as VoIP, with the user population. When these users get better QoS, they are likely to use more services from the network.
For example, a study carried out by Cisco Systems in 2002 concluded that, "on average, end-users say that wireless LAN (WLAN) allows them to be connected 1,75 hours more per day. This translates into an estimated 70 minutes increase in productivity. Taking the productive connection time of 5,5 hrs/day as an average, the advent of WLAN allows an enterprise user to be as much as 22% more productive". Given a fully loaded average yearly salary of $96k, this indicates that the annual productivity improvement is worth, on average, $21k/year/employee. A medium-sized corporate organisation with 200 employees already achieves improvements worth as much as $4,2m per annum.
In the same study, 63% of end-users reported that "WLAN technology improves the accuracy of everyday tasks - with 51% of healthcare organisations finding significant improvements in accuracy. This is a crucial improvement, given the life and death implications at the point of care".
Therefore, it is vital for the enterprise to ensure the timely and qualitative introduction, management and troubleshooting of innovative IT services and applications. To be successful, an ROI calculation should consider business-specific criteria to measure QoS.
Using call centres as an example, improvements in QoS are typically observed in the following operational changes:
* A reduction in overall workforce fluctuation.
* A reduction in the number of truck rolls.
* A decrease in the time spent locating and fixing problems.
* A decrease in the average call duration, due to easier network navigation and faster access to the relevant troubleshooting information.
* A decrease in the number of repeat trouble-tickets.
A well-designed centralised IP-network analysis system will exhibit measurable impact in all of these metrics.
The actual size and the speed of impact on enterprise revenue resulting from a centralised IP-network analysis system also depends on macro-factors such as economic growth. However, for purposes of this white paper, and based on experiences collected at major telecom service provider sites, it is estimated that the quality-enhancing effects of next-generation centralised IP-network analysis systems lead to a 3%-5% growth in revenue.
Improved cash positions
Apart from contributing to an increase in revenue, centralised IP-network analysis systems can also lower operating costs by augmenting the productivity of both workforce and capital. The capability to analyse the network and feed complex processes such as network fault-, configuration-, and performance-management with a limited level of labour and equipment translates directly into reduced costs and ultimately into profit.
This section examines the cost savings associated with centralised IP-network analysis system under two factors of production:
* OPEX (operating expenditure).
* CAPEX (capital expenditure).
Because virtually every dollar of OPEX reduction drops straight into free cash flow, we will start with the OPEX section. On the other hand, estimates show that only about 20c of each dollar of revenue translates into free cash flow.
Improve cash positions: OPEX
There is wide agreement that the labour cost savings associated with centralised IP-network analysis systems are substantial. As stated earlier, a well-planned centralised IP-network analysis system can help achieve operational excellence in various different ways, including reducing labour costs by automating analysis processes, and by reducing transfer time.
For example, in 2002 a large communication service provider completed a project that enables it to carry 100% of telephone calls between the two major cities in its home country, plus 50% of all its international European telephone calls, over its MPLS-based IP network. This represents over three billion minutes of telephone calls per year running over an IP network based on equipment from Cisco Systems.
The following table summarises the labour advantages in the operation and management department. The savings result from the implementation of a centralised IP-network analysis system, instead of dispatched analysis tools deployed on the old circuit-switched network used at that incumbent communication service provider.
The bottom line is a staggering 40% incremental engineering efficiency growth achieved at the featured communication service provider. Since incremental efficiency growth ratios depend on the workforce category, a proper ROI calculation has to consider categories that correspond to the major functions performed by the enterprise.
For example:
Service and application planning and deployment
* Improved SLA management.
Network and service installation
* Faster provisioning of end-to-end circuit paths.
* Faster testing.
* Realtime network statistics available online.
* Better tracking of service provisioning.
Network management and maintenance
* Better tracking of network utilisation.
* Faster resolution of trouble-tickets due to aggregated network-wide view.
* Less reliance on the experience of individuals.
* Better communication through online/offline reporting.
* Reduction in the number of truck rolls.
Customer support and administration services
* Faster response times.
* Fewer repeated calls.
* Reduction in average call duration.
* Increased end-user satisfaction.
In summary, by projecting the above findings on an average enterprise for the purposes of ROI estimates, a centralised IP-network analysis system can lead to an estimated savings in the region of 40% for network installation and maintenance staff.
Improve cash position: CAPEX
Enterprise networks are capital-intensive investments. Large enterprises must invest in expensive network equipment and keep track of valuable network inventories. A lean and homogenous approach is required to make the best possible use of their network resources.
A centralised IP-network analysis system fits well in this space, for it provides an aggregated, network-wide view of the network performance and utilisation, with a single intuitive interface across all LAN/WAN technologies.
Significant cost savings from a centralised IP-network analysis system stem from improved utilisation of the enterprise's existing infrastructure, specifically network, circuits and other assets.
In other words, a state-of-the-art centralised IP-network analysis system affects costs by reducing capital investment through better use of existing network bandwidth. The capital cost savings are greatest in the LAN equipment area since large enterprises typically solve network capacity problems by throwing bandwidth at it. New devices and bandwidth are allocated for customers even if, in reality, existing network capacity is available. Clearly this does not help to solve the root cause of the networking problem.
The impact of the size and speed of a centralised IP-network analysis systems impact on material cost depends ultimately on the specific market segment and growth rate of the enterprise. Findings from independent studies on material CAPEX savings suggest a reduction in the region of 38% on IP Telephony versus circuit switched telephony, for instance.
Apart from material costs, enterprises also incur significant training costs.
A centralised IP-network analysis system can also have a relevant impact on these costs. Here is a case taken from a large, US-based corporation in the media and entertainment segment:
By using configurable network analyser devices and data from network equipment agents that support SNMP, MIB and RMON, a centralised IP-network analysis system can show the overall health of a network. By using the same, common GUI across any LAN/WAN network technology and agents, a centralised IP-network analysis system reduces training requirements de facto. By adding the incremental engineering efficiency to the reduced training requirements stated above, for the purpose of ROI estimation, the financial impact on the training budget in the above example is, again, a reduction of around 40%.
Implementation costs
Although it may be agreed that centralised IP-network analysis systems can help large enterprises to increase revenues and decrease costs, the initial investment must be considered. The main cost components associated with the purchase of a next-generation centralised IP-network analysis system are:
* Initial non-recurring acquisition costs.
* Commissioning costs such as training and data migration from existing tools.
* Eventual integration costs such as integrating the centralised IP-network analysis system with other in-house systems (billing, network management systems).
* Hardware costs such as new distributed data capture devices may be required.
An effective ROI model has to take all of these expenses into account.
How can Agilent help?
For a medium-sized enterprise, and for the purpose of ROI estimation, the non-recurring acquisition costs for the software licenses of Agilent's version of a centralised IP-network analysis system - the Network Troubleshooting Center - is in the region of $30 000, while hardware acquisition costs for Agilent's high-performance Network Analyzer acquisition devices typically start at $20 000. The actual costs depend on the number of devices and interfaces that the centralised IP-network analysis system will be expected to monitor simultaneously. Note that an investment in Agilent's solution is outright ownership, and no re-licensing is required. This increases ROI over a period of time by avoiding recurring additional licensing costs.
Conclusion
By adopting an ROI calculation, large enterprises can benefit from a model that enables them to concentrate the decision of investing in a centralised IP-network analysis system on revenue and cost impacts. By tailoring the data according to their situations, the ROI model can support the calculation of revenue, cost impacts and payback time for any given project.
For more information contact Concilium Technologies, Steve Alves, 012 678 9200, [email protected]
| Tel: | +27 12 678 9200 |
| Email: | [email protected] |
| www: | www.concilium.co.za/test-measurement |
| Articles: | More information and articles about Concilium Technologies |
© Technews Publishing (Pty) Ltd | All Rights Reserved
printer friendly version