A Basic ROI Model


The basic ROI model has two elements: building the case and quantifying the benefits. Often, the process for projecting the costs and benefits involves educated guesswork, based on multiple sources, and just plain guesswork. The assessment process can be very effectiveif the appropriate data have been collected over the evaluation period.

As shown in Figure 13-1, the goal is to determine when the project has broken evendelivering additional value that matches the costs. The project begins in the redthere have been expenses with no recouping of the investment yet. The starting point coordinates are determined by the purchase and implementation costs (vertical axis) and the deployment date. As the investment becomes operational, it begins generating business value, such as increased revenue, reduced costs, or higher service quality.

Figure 13-1. Return on Investment Projections


At some point in time, the cumulative benefit value matches the costs and the break-even point is reached. The Time to Value is frequently used to describe the time needed to reach the break-even point; shortening this time interval recovers the costs more quickly and increases the leverage gained from that investment. The slope of the operations curve indicates the rate of recoverya steeper slope is a shorter Time to Value and a higher payback for each succeeding interval.

Each project alternative has its own ROI graph to find its Time to Value and value slopes. For example, the solid line in Figure 13-1 indicates a stronger ROI potential than the dashed one because the solid line's Time to Value is shorter and its value grows more rapidly over the same time interval. The investment continues to provide additional value after reaching the break-even point. Cost savings and deferred spending are direct business benefits that can be realized throughout a long operational life.

The two ROI lines in Figure 13-1 are linear for simplicity; actual projections may be curved or have discontinuities because of variations introduced by seasonal behavior, sudden market shifts, or other events.

ROI graphs for long-duration projects should include calculation of net present value (NPV). Cash today is worth more than cash tomorrow because cash can be invested and earn interest or another type of return. Therefore, if $1000 is invested in Web systems today to obtain $1000 of benefits a year from now, the project is in the red. That $1000 could have been invested in the financial markets and would have returned more than $1000 over the year.

NPV calculations or similar calculations (such as internal rate of return [IRR]) are used by financial officers to handle this analysis. They use the time value of money, which is the return that money can earn if invested, to make the benefit of an investment clear. After all, investing the money carries much less risk than using that money to improve a Web system. If the money can earn more through investment than it will bring in if used to make those Web system improvements, the project may not be worth the effort.

The ROI Mission Statement

Any ROI project needs a goal as a starting point. In other words, what problem will this investment solve? This mission statement sets the perspective as details are fleshed out for the process. Specific metrics and measurement procedures are determined after the problem is identified. The metrics and measurements will be used to assess the actual ROI generated. In addition, the appropriate baseline measurements can be applied to establish the actual service conditions prior to deployment.

A load-balancing investment, for example, could be assessed by measuring the impact of the investment on some combination of operational metrics, such as the following:

  • Transaction delay

  • Average server loading

  • Number of concurrent connections and transactions supported with acceptable performance

  • Reduced number of reported problems caused by overloaded or down servers

  • Reduced support staff workload

An investment in content delivery infrastructure is another example. The metrics that might be used to assess the ROI potential for content delivery investments (as discussed in the "Content Distribution" and "Instrumentation of the Server Infrastructure" sections of Chapter 9, "Managing the Server Infrastructure") are as follows:

  • The projected, or measured, bandwidth gain at the network edge. Bandwidth gain is used to estimate the bandwidth savings, relative to a centralized distribution system, when a content distribution network is used.

  • Download time

  • Availability

Project Costs

The costs and the time of deployment locate the starting point in Figure 13-1. There are a number of factors to consider, and each project will need to select and weigh the factors that apply to that specific evaluation.

Many factors can be incorporated into a cost calculation, including the following:

  • Staff time for design, project evaluation, project implementation, and project management

  • Consulting time, when needed

  • Training

  • Product costs (hardware, software, services, and management systems)

  • Ongoing maintenance and licensing fees

Some of these elements may be harder to identify and quantify, depending on the specific organization. For example, the requirements may come out of a planning group that addresses a range of technology issues. Time spent on evaluations may include actual testing scenarios, specification reviews, and customer research. The implementation costs might include some vendor-provided professional services, services from other sources, training for operations staff, modifications and updates to management tools, or additional hardware for parallel operation and gradual cut-over.

The cost of maintaining the status quo must also be considered. For example, consider a key service that generates $12 million in annual revenue. If the average availability is 98 percent, there is a potential $240,000 revenue loss (every month the company loses another $20,000 in potential revenues). Investing $50,000 to raise availability to 99.5 percent is an attractive option because the improvement generates $180,000 annually in new revenue opportunities. The Time to Value is less than four months, and each month past the break-even point adds another $15,000 (potentially) to the revenue stream.

At the same time, calculating the costs of the contributors and the alternatives is not always black and white. For example, will all the headcount involved in development and deployment be involved in the project 100 percent of their time? Does an uptime of 98 percent mean that revenue grinds to an absolute halt when the systems are down, or could there be alternatives that keep the revenue going? For example, orders phoned in or faxed in instead of submitted over the Internet are still orders. It's useful to apply a little skepticism to estimating both costs and benefits, as you'll see in the following sections.

Project Benefits

Estimating the benefits from the investment may involve various levels of guessing and using rules of thumb. One of the most frustrating aspects is trying to find some estimates that at least bring you closer to understanding the costs and benefits of an investment. One major question is this: How reasonable are the estimates? Is a 30 percent improvement in bandwidth utilization within the realm of possibility, or is 15 percent more realistic? Each estimate changes the slope of the ROI line and shifts the Time to Value.

Many vendors now have ROI calculators that they use as part of the sales process. They generally embed a set of assumptions about costs and impacts within their calculator. Of course, the estimates are helpful only to the degree that they match the environment. The size of the company, the industry segment, and the relative technical maturity are among the factors influencing the ROI outcomes. Obtaining more information on the embedded information, how it was gathered, and the data used to build the model will help calibrate the results.

I have worked with several vendors to begin building some rules of thumb from early adopter experiences. The goal is to get some idea of what other customers report about their experience with implementation, introduction, and day-to-day operations. These rules of thumb will give potential buyers some additional ways to relate to the usual ROI information. For example, a potential customer can relate to a rule of thumb that indicates that similar customers have reduced their staffing levels by 20 percent while improving service quality. (Some of the rules that can be applied are described in the following subsections.)

Measurements should be analyzed to determine the actual benefits derived from the investment. The key is having sound measurements and a starting baseline. Care should be taken to document performance levels fully and quantitatively before implementation begins. As the implementation proceeds, it is also important to remain flexible and incorporate measurements for tracking unexpected outcomes. Following up with the actual assessment is very helpful for the IT group. It helps them calibrate their own estimates and refine their ability to project benefits with more accuracy in the future. Equally important is establishing their credibility with the business managers through accurate projections. Remember that a business-knowledgeable IT manager should be involved in the analysis to improve the relevance for other business managers.

Availability Benefits

Changes in availability are usually evaluated in terms of capturing more revenues or distributing more information to consumers. Identifying the actual revenue rates is often the most difficult step. Good service instrumentation is necessary to determine the number of orders and the revenue generated when the system is available. In some cases, however, a simple calculationdividing the revenues by the number of operational hours to get an average revenue ratemay suffice.

After the revenue rate is determined, the rule of thumb is that a 1-percent change in availability is an additional 7.2 hours per month of potential revenue gain. If your organization has annual revenues of $12 million, for example, the rate is $1 million monthly, or almost $1400 per hour. A 1-percent change can produce an additional $10,000 in monthly revenues, or $120,000 on an annual basis. Spending $50,000 to raise availability by 1 percent gives an ROI Time to Value of five months.

Performance Benefits

Performance is easy to measure for compliance: it is straightforward to determine the percentage of transactions that complete within the specified response time. Assessing the business impact is a bit more difficult. A subset of the total transactions actually generates revenues; the remaining transactions are used for browsing product information, checking promotions, facilitating customer-managed support, or tracking outstanding orders, among other possibilities. Identifying the types of transactions and tracking each category may be required.

Tracking the actual business that is transacted and closed provides deeper insight into the impacts of any investment. Some metrics that demonstrate the value include the following:

  • Change in the completion rate How many potential revenue-generating transactions completed successfully?

  • Change in the deal size Are the orders larger as a result of better service quality?

  • Change in unit transaction costs The investment may allow larger transaction volumes without requiring expenditures for additional infrastructure, thereby reducing the unit costs of transactions.

Staffing Benefits

Staffing impacts may be important considerations in some situations. For example, investments in automated tools may allow staff head count reductions or reassignments to other tasks. The savings include salaries, benefits, and possibly training costs. At other times, staffing may remain constant while the infrastructure grows over time, improving staff productivity by managing more elements and services without adding team members. In the end, the best metric for services management is the change in transaction volumes relative to staffing levels.

Infrastructure Benefits

Many investments will impact one or more infrastructures. A basic metric is the change in service flows relative to the infrastructure changes. If an infrastructure handles more service flows as a result of the investment, its productivity has been improved accordingly. For example, a company I interviewed was able to increase their transaction flows by 30 percent without additional spending, translating into a substantial savings compared to scaling the infrastructure by that amount.

Deployment Benefits

Deploying new services is an ongoing process rather than one of periodic releases. Rapid deployment is facilitated with load testing and good design practices. Load testing also helps determine if a new service will degrade the current services mix when it is placed into production. The usual response is to refuse to deploy any new services that disturb the normal operational baselines.

Deployment delays can have significant impacts on revenuesa rule of thumb that I use is a $20,000 revenue loss for every $1 million in annual revenues that the service generates. Indirect impacts include disappointing customers and losing competitive advantage to other early movers. Using predeployment load testing and other service-level management techniques can decrease the probability of deployment delays and, therefore, improve revenues.




Practical Service Level Management. Delivering High-Quality Web-Based Services
Practical Service Level Management: Delivering High-Quality Web-Based Services
ISBN: 158705079X
EAN: 2147483647
Year: 2003
Pages: 128

flylib.com © 2008-2017.
If you may any questions please contact us: flylib@qtcs.net