Slow performing networks and applications – How do you quantify the business impact?

By Fluke Networks

Enterprises of all types and sizes – across a wide array of vertical markets – are increasingly relying on the network and the applications that traverse it for organizational efficiency and business productivity.

In hard economical climates, like we are in right now, it is more important than ever to get the most out of each single dollar spent in IT while still optimizing service delivery of business critical applications to the end user. Delivery of poor performing services can have up to 9% revenue impact depending on vertical industry. Complete service failures not taken into account! It seems that enterprises today can "see" the value of improving network application integrity and efficiency via proactive monitor­ing and management of applications and the infrastructure. However, the financial impact on the business is still a very unclear factor. The same is true when it comes to "proving" the value of improved performance and efficiency by quantifying savings or reduced costs through a positive return on investment.

There is a huge difference between qualitative versus quantitative assessments. Typically, individuals within an enterprise focus on the qualitative point of view instead of quantifying the impacts of improved services and reduction in financial risks or exposure. It is virtually impossible to calculate a return on investment - either positive or negative - with qualitative assessments only. This article focuses on how to quantify the potential savings gained by enhancing network and application performance management.

Improving network application integrity, performance and efficiency

An enterprise's reliance on its network and applications is likely to continue growing exponentially. Revenues, compliance, inventories, and customer support are inextricably linked with network and application performance. While the business-critical applications running over the network vary by enterprises - ranging from SAP and Oracle to Voice over IP (VoIP) and home-grown applications - the exposure of poor performance can mean bottom-line impact ranging from the thou­sands to the millions of dollars.

To understand how the IT infrastructure impacts both applications and users, IT managers require the effective, flexible tools that can provide the highest degree of visibility across the infrastructure. The objective of the system is to enable IT departments to measure the success in business terms based on the high performance of the mission-critical applications.

Quantifiable impacts of better network efficiency

When quantifying the impact of improved application integrity and network efficiency, there are generally two types of savings - hard costs and soft costs. Hard-cost savings are tied directly to the amount of money saved from the impact. Soft-cost savings are typically tied to improved staff productivity so employees are more efficient and can do more with their time.

Areas of impact

Increasing revenue and production capabilities

With the increased importance of the network and the applications traversing it, every enterprise must be able to answer the following question, "If my most business-critical application is down for a period of time, what is the cost per hour to my organization?" These costs can vary quite a bit depending on the enterprise. A financial institution might lose $500,000 in transactions for every hour of downtime, a manufacturing company might lose $250,000 by having to shut down a production line because the just-in-time inventory system was taken offline. Each of these scenarios has three components - the area of the business impacted, the financial risk, and the duration of the incident. Improving network application integrity and efficiency only reduces the amount of time the business is impacted, but the value is immense.

Optimizing bandwidth expenditures

The cost of bandwidth can consume as much as two-thirds of total networking budgets for many enterprises. IT organizations must walk a fine line between having sufficient bandwidth resources for business-critical applications and not wasting critical budget dollars by over engineering the network infrastructure.

Compounding the issue for many enterprises is the proliferation of bandwidth-consuming applications across the enterprise. Enterprises are rolling out new applications that are impacting existing networks including VoIP, Oracle, SAP and Citrix. When enterprises deploy new applications, many times they add bandwidth not knowing if and where they actually need the increased resources. With tighter IT budgets, the days of throwing bandwidth at issues such as these are numbered.

Upgrades, only when necessary

A medium-sized organization is planning on rolling out VoIP across its 22 domestic locations. Due to the perceived utilization of the existing network and the delay-sensitive nature of VoIP, the enterprise decided to increase the speed of each location. (18 sites by 256Kbps, 4 other larger sites by doubling the T1). The total annual cost of the upgrade would be $126,240.

With a tool that provides detailed utilization, an enterprise can "right size" the circuits - having sufficient bandwidth for critical, delay-sensitive applications without wasting resources. Of the 18 sites that were getting an upgrade, only six really needed it. Of the four larger sites, one needed the T1 upgrade. So instead of spending $126,240 annually on upgraded bandwidth, the enterprise only needs to spend $59,040. That is a cost savings of $67,200 or a reduction of 53.4%.

Improving identification and repair of virus and worm attacks

Even though every enterprise has firewalls, virtually every organization has been penetrated to some degree within the past year by a virus or worm attack. Enterprises tend to classify virus attack impacts in two ways: severe threat and residual impact. Severe threats degrade the applications and network to such an extent that key programs or business activities are impacted; leading to possible lost revenue or incurred costs. The residual impact occurs once the severe threat is over but the enterprise must still update its systems and perform mainte­nance. On average, critical threats tend to last approximately 3-6 hours while residual impacts tend to last several days.

The cost of a virus attack

Over the past year, this enterprise has experienced five virus/worm attacks with a varying degree of pain. The severe threat length ranged from two hours for a smaller, common security breach to nine hours for a severe attack that crippled the organization for over a day. For this enterprise, the cost of application and network downtime/degradation has been calculated to $9,000 per hour (this amount is impacted by criteria such as the size of the organization, the number of users and the importance of application and network integrity).

Determining a savings for improved virus/worm detection and resolution is based on how quickly the attacks can be solved compared to today's averages. With the network application performance solution, the amount of downtime/degradation should be reduced 70% by identifying the spike caused by the attack and using the troubleshooting tools to isolate and solve the problem. In this scenario, the savings is calculated to be $63,000 (1.4 hours saved x 5 events x $9,000 per event).

Reducing application MTTR

For most organizations, poor application performance is likely the most difficult and troubling problem. There are so many factors that can cause poor application performance, ranging from the local loop to the circuit to the network to the application servers. And if the problem is intermittent, the challenge grows exponentially.

The key to reducing application MTTR is the ability to isolate where - across the wide spectrum of possible causes - the problem is occurring. More important is the ability to go back in time, whether an hour, a few days or a couple of weeks, to easily identify what caused the intermittent problem so it does not return and pose greater threats later. The goal for enterprises should include identifying degradation in application performance before end users are impacted, thereby reducing the risk to the organization.

Solving the hard-to-isolate problem application

This calculation deals with typical day-to-day issues in identifying and resolving trouble tickets like "I can't process the order" or "I can't get to e-mail." For this enterprise, problem identification and resolution are harder because it isn't a single catastrophic issue causing the performance issue. For these types of intermittent issues, the IT organization focuses on when users cannot use the applications completely and estimates approximately six hours to resolve these issues. This does not include poor performance where applications may be slow but can still get through. The cost of poor application performance is $6,000/hour for this enterprise.

By having a robust troubleshooting tool which provides extensive Layer 1-7 visibility and both a real-time and a graphical "back-in-time" view, the amount of time saved in solving the problem has conservatively been estimated at 50%. For the savings calculation, the enterprise experiences this severe impact once every two months. The savings would be $108,000 annually for this scenario by reducing application MTTR (3 hours saved x $6,000/hour x 6 impacts).

Improving network user productivity

A challenge for many enterprises is quantifying the impact of poor application and network performance for end users. While it is simple to say, for example, end users will be impacted if the order processing system is down for 6 hours, it is much harder to precisely quantify the financial impact. There is a substantial difference between a stockbroker who cannot place a trade and a marketing person who may be inconvenienced, but can work on other tasks. The critical step with quantifying the impact of poor network productivity is providing a realistic view of the impact of application and network downtime and degradation.

Employees require access to business applications

In this scenario, there are many variables tied to the calculation of the savings. This enterprise has 2,500 end users with an average weighted salary of $37,500 ($25,000/year + 50% benefit cost). Historically, the enterprise determines application and network degradation impact performance substantially approximately 2% of the time. Of the 2,500 users, approximately 10% will be negatively impacted for business critical activities (logistics and production are impacted, but sales, marketing, finance, HR, etc. are not affected).

With enhanced visibility and troubleshooting of network and application performance management, the amount of downtime is estimated to be reduced 45%. By reducing application downtime, the enterprise would save $484,375 in network user produc­tivity (2,500 users x $37,500 loaded salary x 2% downtime x 10% end user impact x 45% savings).

Calculating a payback period and ROI

When deploying a new solution, enterprises today typically look at a payback period (when the savings pays for the solution's savings pay for the cost of deployment) or return on investment (ROI). Once you quantify the savings, you can calculate a payback period and a return on investment (ROI). The payback period determines the amount of time it takes for the performance management system to pay for itself. A payback is calculated by dividing the cost of deployment by the savings and multiplying by 12 months.

A ROI quantifies how much return you will receive on an investment over a period of time. Most ROIs are calculated in 1-, 2- or 3-year projections. The ROI is calculated by dividing the total yearly savings by the cost of deployment.

Armed with this quantitative analysis, your enterprise can make an informed decision on the benefits and savings associ­ated with a proactive, application and network performance management system. You can run different scenarios using the same toolset to be more aggressive or more conservative and see how the payback period and ROI change.

Conclusion

Believing there are benefits to improved network and application performance is not enough for most organizations. The ability to quantify cost savings, improved productivity or reduced risks is a critical component in justifying an investment in application, VoIP and network performance. This paper has listed some key areas where cost savings can be quantified. Each organization will have different results and savings - some savings might be spread out evenly while others will be skewed to only one or two criteria.

Contact details:
Benny Vogels
EMEA Marketing Manager, Performance Solutions
T: +49 6103 3012644, E: benny.vogels@flukenetworks.com