IT Utility

There are literally hundreds of business processes taking place simultaneously in an organization, each creating value in some way. The art of strategy is to identify and excel at the critical few processes that are the most important to the customer value proposition.

Both private companies and governmental agencies have outsourced some of their computer processing systems to third parties. Processes commonly outsourced include

1. Asset management
2. Help desk
3. Infrastructure maintenance
4. Systems management and administration
5. Network management
6. Integration and configuration

These outsourced information technology (IT) services have come to be known as the “IT utility.” The larger IT utilities are typically ISO 9001/9002 certified and offer large pools of IT talent and experience. However, processes must be measured, regardless of whether or not they are outsourced.

Unisys (2003), a provider of such services, recommends the following metrics:

1. Customer satisfaction
2. Standardization
3. Incident rates
4. Security audit
5. Incident prevention rates
6. Security awareness
7. Availability
8. Reliability/quality of service
9. Call volume
10. First pass yields
11. Cycle times
12. Architecture accuracy
13. IT employee satisfaction
14. Root-cause analysis
15. Change modification cycle times
16. Change modification volume by type
17. Research and development (R&D) presentation/information flow rate
18. Volume of technology pilots
19. Business opportunity generation rate
20. Strategic IT project counts

Unisys uses these metrics to establish the foundation for management review, trend analysis, and causal analysis. Management review provides insight into current performance and forms the basis for taking corrective action. Trend and root-cause analyses identify opportunities for continuous improvement.

Based on its analysis and industry experience, Unisys states that a performance-based environment is anywhere from 10% to 40% more cost-effective than a non-performance-based environment. When deciding how best to optimize the IT infrastructure, organizations need verifiable performance and trend data. While customer satisfaction is usually touted as the key metric for IT improvement, it is actually an outcome metric dependent on several lower-level activities, as shown in Figure 8.1. Understanding the relationship between these codependent performance metrics is important in effecting sustainable positive performance.

Essentially, the IT department should consider itself an IT utility for the purposes of aligning itself to the organization’s business process objectives. In this way, IT can more effectively track performance using a balanced scorecard and make appropriate performance improvements as a result.

Wright et al. (1999) did an extrapolation of Compaq’s (now Hewlett-Packard) balanced scorecard based on research and publicly available information. As a computer company, this case history is interesting from an IT perspective, particularly if a corporate IT department thinks of itself as an IT utility.

Compaq had a number of business process objectives:

Figure 8.1 Cause and effect in the IT utility.

The business process perspective is linked downward to the learning and growth perspective by quality improvements, improved coordination, and integrated information. The business process perspective is linked upward to the customer and financial perspectives by lower operating costs, improved use of resources, reduced waste, new product capabilities, and better service programs.

For Compaq, the chief component of the business process perspective is its operations cycle. This encompasses sourcing parts, components, manufacturing, marketing, distributing, and after-sale services. This cycle had been the major focus of a reengineering effort—the goal of which was to bring them to a higher level of customer focus.

Compaq’s reengineering effort required them to change their business processes first and then their information systems to support the reengineered processes. The company relied heavily on enterprise-level IT, and by the late 1990s had begun using SAP R/3 to integrate their business processes and sales information. Compaq also built a global extranet called Compaq On Line to provide customers with a way to automatically configure and order PCs and servers. This was followed by adding an online shopping service, allowing customers to order directly from the Internet.

The newly enhanced processes and accompanying systems allowed Compaq to achieve the following process efficiencies:

1. Linking orders electronically to suppliers. This improved cycle time and facilitated JIT manufacturing. It also provided production status information to be made available to customers so that they could track their own orders.
2. Sharing information with suppliers enabled Compaq to anticipate changes in demand and ultimately improve their efficiency. This reduced the cost of supplies and improved on-time delivery.
3. Integrating orders with SAP’s financial management and production planning modules enabled Compaq to reduce the time and cost of orders.
4. Capturing customer information after a sale enabled Compaq to provide individualized service and additional marketing opportunities.

After the implementation of a balanced scorecard in 1997, Compaq sales volume improved. According to Wright et al. (1999), this resulted from delivering value, increasing customer service, innovating new products, and reducing time-to-market (TTM). This sales spurt more than made up for the decreasing prices of PCs, and ultimately generated higher revenue. Improved cycle times and decreasing costs enabled the company to operate far more efficiently, resulting in higher net income levels and, ultimately, higher revenue per employee.

Of course, a balanced scorecard is not the only performance-enhancing methodology and measurement system in use. Aside from Six Sigma, which is a set of techniques and tools for process improvement developed by Motorola in the 1980s, deployed methodologies run the gamut from business process management (BPM) to kaizen.

BPM is a field in operations management that focuses on improving corporate performance by managing and optimizing a company’s business processes. Design for Six Sigma (DFSS) is a BPM methodology related to traditional Six Sigma. It is based on the use of statistical tools such as linear regression and enables empirical research.

Define, measure, analyze, improve, and control (DMAIC) refers to a data-driven improvement cycle used for improving, optimizing, and stabilizing business processes and designs. Kaizen is the Japanese term for improvement. Kaizen refers to activities that continuously improve all functions and involve all employees from the chief executive officer (CEO) to the customer service reps. It also applies to processes. Finally, Lean software development (LSD) is a translation of Lean manufacturing and Lean IT principles and practices to the software development domain.

Getting to Process Improvements

Process improvements can be thought of in two dimensions. There are those process improvements that are internal to the IT department and there are process improvements that are quite visible to end users and senior management. For the purposes of this discussion, we will refer to the former as engineering process improvements and the latter as business process improvements.

Table 8.1 defines five levels in terms of continuous improvement.

Enhancing IT Processes

The Boston Consulting Group (Brock et al. 2015) report on large-scale IT projects paints an ugly picture. For projects over $10 million in investment, the chances of delivering on time and on budget is just one in ten. Projects fail for a myriad of reasons: chief among them is overly complex requirements, inexperienced teams, a lack of buy-in from influential stakeholders, insufficient attention to major risks, lack of testing, and a failure to manage the project plan effectively.

A report by the Hackett Group (2015) stresses that chief information officers (CIOs) must respond far more quickly to shifts in business demands, while decreasing expenses through greater efficiencies. They recommend that CIOs overhaul the service delivery model to better support the convergence of business and tech. The goal here is to introduce greater agility and flexibility so that IT can “turn on a dime” to support the changing needs of the business and create real strategic advantage.

Today’s fluid marketplace requires technology that can drive innovation, automation, and personalization much more quickly (Desmet et al. 2015). As a result, some organizations are moving to a two-speed IT model that enables rapid development of customer-facing programs while evolving core systems designed for stability and high-quality data management more slowly.

This translates to the use of high-speed IT teams that are charged with rapidly iterating software, releasing updates in beta, fixing problems in near real time, then rereleasing. One European bank created a new team that used concurrent design techniques (in which multiple development tasks are completed in parallel) to create a prototype of an account registration process, while using existing technology where it could. By testing this process with real customers in a live environment, the team was able to make constant refinements until it succeeded in cutting the process down to 5 steps from the original 15. The key here is knowing what questions to ask as one goes about improving a process, as shown in Table 8.2.

New Methods

DevOps (the integration of technical development and operations) and continuous delivery (the automation of testing, deployment, and infrastructure processes) have introduced capabilities that radically increase speed to market and lower costs. An international travel company used these approaches to reduce TTM by moving to the cloud, fully automating its testing, and rolling out a one-click deployment process (Desmet et al., 2015).

The concept of continuous development is closely linked to the Lean/agile concept of kaizen-continuous improvement. For some teams, kaizen can lead to continuous delivery and continuous deployment. Intuit’s accounting software runs on a wide variety of platforms, so it is obvious that Intuit’s development team has many technical challenges in maintaining the infrastructure that supports continuous development. Intuit’s method is known as infrastructure as code, a fully automated development, test, and production infrastructure (Denman 2015). Continuous development allows development teams to test every code and commit and deliver the working code as soon as it is ready, with a minimum of manual testing. Essentially, every code gets automated unit tests, smoke tests, and so on to catch any defects.

At this time, only a few of Intuit projects have fully achieved continuous deployment, where changes go into production without any manual intervention. Some projects have continuous delivery, meaning the code is pushed out to a staging environment where it is reviewed and approved before being sent into production. Other projects require manual tests by quality engineers before they can be deployed. But, the ultimate goal is complete automation and full continuous deployment, and infrastructure as code is the software that runs all that.

A push toward continuous development, which includes continuous integration; continuous testing; automated configuration management (CM) and testing; use of agile methods, including use of patterns; and application performance monitoring, needs to be handled with care. There really needs to be a culture shift within the development teams, including project managers, programmers, database designers, web designers, quality assurance (QA), and operations and management, toward enhanced collaboration.

Process Quality

Solano et al. (2003) have developed a model for integrating systematic quality—that is, a balance between product and process effectiveness and efficiency—within systems development organizations through the balanced scorecard. Table 8.3 shows the four balanced scorecard perspectives oriented toward systemic quality integration.

This quality-oriented strategy is a daily, ongoing process that needs to be “bought into” by staff members. Solano et al. (2003) provide an example of a company, VenSoft C.A., which did just this by relating organizational goals to employee remuneration. Table 8.4 shows employee incentives, based on the balanced scorecard. Each perspective and indicator was given a weight that depended on the organization’s mission. Yearly bonuses depended on the goals being totally or partially attained.

How is the process (or product) quality index calculated? One of the goals of software engineering is to produce a defect-free product. A module’s quality profile is the metric used to predict if a module will be defect free. A quality profile is predictive in that its value is known immediately after a module has completed its developer’s unit test. It is suggestive, in that it can suggest potential quality issues and thus mechanisms to redress those issues. Quality profiles adhere to software engineering dogmas that design is good, technical reviews are necessary for quality, and that high-defect density in a test phase is predictive of high-defect density in later test phases. Finally, early empirical evidence suggests that quality profiles do predict if a module is defect free. As seen in Table 8.5, a quality profile is composed of five dimensions:

The process quality index (PQI) is calculated by multiplying the five dimensions together. The Software Engineering Institute (SEI) has presented preliminary data that indicate that PQI values between 0.4 and 1 predict that the module will have zero subsequent defects.

With this model, Solano et al. (2003) tried to close the gap between software engineering projects and organizational strategy. In their view, the systemic vision of the organization, and the balance between the forces of the organization, coincide quite nicely with the balanced scorecard approach.

Philips Electronics (Gumbus and Lyons 2002) implemented a balanced scorecard predicated on the belief that quality should be a central focus of their performance measurement effort. The Philips Electronics balanced scorecard has four levels. The very highest level is the strategy review card, next is the operations review scorecard, the third is the business unit card, and the fourth level is the individual employee card.

The corporate quality department created very specific guidelines for how metrics should link the cascaded scorecards. These guidelines indicate that all top-level scorecard critical success factors (CSFs) for which the department is responsible must link metrically to lower-level cards. Three criteria were established to accomplish this:

1. Inclusion: Top-level CSFs must be addressed by lower-level CSFs to achieve top-level metric goals.
2. Continuity: CSFs must be connected through all levels. Lower-level measurements should not have longer cycle times than higher-level measurements.
3. Robustness: Meeting a lower-level CSF goal must assure that high-level CSF goals will be met or even surpassed.

As you can see, the goals in all card levels align with the goals in the next level above, and the goals become fewer and less complex as you drill down through the organization.

The CSFs, selected by the departments that had a major controlling responsibility, were the key balanced scorecard indicators. The management team of each business unit selected CSFs that would distinguish the business unit from the competition. They used a value map to assist in determining the customer CSFs and then derived the process CSFs by determining how process improvements can deliver customer requirements. Competence CSFs were identified by figuring out what human resource competencies were required to deliver the other three perspectives of the card. Standard financial reporting metrics were used for the financial perspective.

At this point, each business unit was charged with figuring out what key indicators could best measure the CSFs. The business units had to make some assumptions about the relationships between the processes and results to derive performance drivers and targets. These targets were set based on the gap between current performance and what was desired 2 and 4 years into the future. The criteria for these targets were that the targets had to be specific, measurable, realistic, and time phased. The targets themselves were derived from an analysis of market size, customer base, brand equity, innovation capability, and world-class performance.

Indicators selected included

1. Financial: Economic profit realized, income from operations, working capital, operational cash flow, inventory turns
2. Customers: Rank in customer survey, market share, repeat order rate, complaints, brand index
3. Processes: Percentage reduction in process cycle time, number of engineering changes, capacity utilization, order response time, process capability
4. Competence: Leadership competence, percentage of patent-protected turnover, training days per employee, quality improvement team participation

In cascading the scorecard throughout its different levels, six indicators were key for all business units:

1. Profitable revenue growth
2. Customer delight
3. Employee satisfaction
4. Drive to operational excellence
5. Organizational development
6. IT support

In one of the business units, Philips Medical Systems North America, results are tracked in real time. Data are automatically transferred to internal reporting systems and fed into the online balanced scorecard report with the results made immediately accessible to management. The results are then shared with employees using an online reporting system they call Business Balanced Scorecard On-Line. To share metrics with employees, they use an easy-to-understand traffic-light reporting system. Green indicates that the target was met, yellow indicates inline performance, and red warns that performance is not up to par.

Process Performance Metrics

Some researchers contend that organizations are shooting themselves in the foot by ignoring web analytics. Swamy (2002) states that without this link, a major portion of the organization’s contributions to success and/or failure is missing. He contends that most online initiatives have a dramatic impact on offline performance. Therefore, excluding web analytics, as immature as these statistics are, precludes senior executives from seeing the whole picture.

Figure 8.2 Web analytics added to the balanced scorecard.

Swamy recommends adding two new perspectives to the balanced scorecard, as shown in Figure 8.2.

IT processes are project oriented. Stewart (2001) makes the following metric recommendations when establishing project-specific balanced scorecards:

• Financial
  1. On time
  2. Within budget
  3. Variance from original baselined budget and final budget
  4. Project costs as compared with industry standard and organizational standards for similar projects
  5. Earned value

• Customer
  1. Project meeting intended objectives
  2. Customer satisfaction (including account payment history)
  3. Economic value added (strategic benefits rather than financial benefits achieved referability, increased venture capital support, etc.)

• Project/internal business
  1. Project resource requirements management
     1. Average management time of project manager related to total effort

  2. Project portfolio comparatives
     1. Project cancellation rate
     2. Project backlog—awaiting start-up
     3. Risk management statistics
     4. Contingency time allotted and used

  3. Change management statistics (number of change records per designated period of time can show whether proper project scope has been set, percentage change to customer/vendor environment impact to scope)
  4. Quality management statistics (rework, issues, etc.)
  5. Project team member satisfaction

• Growth and innovation
  1. Average capabilities per team member and improvement over course of project
  2. Development or ongoing improvement of templates, procedures, tools, and so on
  3. The rate that innovative ideas are developed (new ways of doing things)
  4. Best practices identified
  5. Lessons learned and applied
  6. Positive achievements/impacts to the organization
  7. Evaluate quantitative statistics
     1. Examine true costs of operation, evaluating impact of project slippage and inadequate support and nonsupport infrastructure costs

  8. Evaluate organizational change
     1. Evaluate how the change has impacted the organization’s business

  9. Reduce to lowest common denominator
     1. Review support costs versus costs of delay per person
     2. Review actual project costs versus plans (net present value)

  10. Review strategic objectives achieved
      1. Review qualitative statistics
      2. Identify unanticipated benefits accrued
      3. Review attainment or contribution to organizational objectives versus time commitment

  11. Review overall business value improvement
      1. Revenue increase/decrease
      2. Team retention and promotion
      3. Increased market share, references

Shared First

The goal of the “shared-first,” which is an approach recommended by the U.S. federal government’s CIO Council (2013) is to improve performance, increase return on investment (ROI), and promote innovation. Through shared services, separate departments within an organization may eliminate duplication of cost structures, reduce risk, procure needed services, implement new capabilities, and innovate in a rapid and cost-efficient manner. The specific goals for implementing shared services include

1. Improve the ROI through the coordinated use of approved interdepartmental shared services
2. Close productivity gaps by implementing integrated governance processes and innovative shared service solutions
3. Increase communications with stakeholders as managing partners, customers, and shared service providers work together to ensure value for quality services delivered, accountability, and ongoing collaboration in the full life cycle of interdepartmental shared services activities

“Shared first” principles will produce a number of beneficial outcomes, which include the following:

1. Eliminate inefficient spending that results from duplicative service offerings and systems
2. Enhance awareness and adoption of available shared services across the organization; promote agility and innovation by improving speed, flexibility, and responsiveness to provisioning services through a “shared-first” approach
3. Focus more resources on core mission requirements rather than administrative support services
4. Spur the adoption of best practices and best-in-class ideas and innovations
5. Reduce the support costs of redundant IT resources
6. Improve cost-efficiencies through shared commodity IT

There are two general categories of shared services: commodity IT and support. These may be delivered through cloud-based or other shared platforms. Commodity IT shared services opportunities include IT infrastructure (e.g., data centers, networks, workstations, laptops, software applications, and mobile devices), and enterprise IT services (e.g., e-mail, web infrastructure, collaboration tools, security, identity, and access management). Commodity IT is asset oriented, while enterprise IT services may, at times, be more utility oriented (defined as purchasing-by-usage rate). Support services are defined by the capabilities that support common business functions. These include functional areas such as budgeting, financial, human resources, asset, and property and acquisition management.

The following steps indicate tasks and activities, best practices, and risk areas with mitigations to consider and prepare for when implementing shared services.

Configuration Management

CM provides the means to manage technology-related processes in a structured, orderly, and productive manner. As an engineering discipline, CM provides a level of support, control, and service to the organization. CM is a support function in that it supports the program, the corporation, and, in a number of situations, the customer.

The process of CM has not really changed much during the past 20–30 years. However, the environment that CM operates within has changed significantly and is likely to continue to change. Regardless of the fast-paced changes in the technology arena, the process of CM is basically immutable—that is, the process does not change, only what is being managed changes. The key is in the process.

In Conclusion

Organizations are composed of systems, which are composed of individual processes. For an organization to improve its performance, it must embark on a systematic approach to optimize its underlying processes to achieve more efficient results.