Good news! Software Change Management: Case Studies and Practical Advice, by Donald J. Reifer, is now available for purchase.
Donald J. Reifer is one of the leading figures in the field of systems/software engineering and management, with over 40 years of progressive experience. He has built businesses, steered troubled projects, and served in executive positions in industry and government. He specializes in the area of metrics and measurement. Besides being a trusted advisor to major corporations and the government, he has founded software businesses and served as an expert witness. Check out Donald’s recent post to the Microsoft Press blog.
The list of contents and some additional information from the Introduction can be found in this previous post.
If this sounds interesting to you, take a moment to read this excerpt from Chapter 6, “Industrial Case: Moving to the Clouds.”
The utility company you work for is considering moving to cloud computing. Cloud computing, as you find out via the Internet, is a relatively new concept that refers to an assortment of logical computational resources that are made available via computer networks rather than on local computers. 1,2 Applications are hosted on multiple servers across the cloud. Data is also stored on server farms. In this manner, both applications and data can be made accessible via a browser rather than you having to install and run them on your desktop, laptop, or office server. Instead, both run on the cloud via its servers and results are made available through the network to clients on their computers. Clouds can be public and private, depending on the need. In addition, cloud services are sold on a demand basis using any of these three arrangements:
■ Software as a service End-user applications services are accessed over the network rather than on client computers. Under this arrangement, you execute your business application remotely to get results typically at a fraction of the cost of licensing the software.
■ Platform as a service Sets of application components can be put together by developers via plug-and-play and run on cloud-computing servers to get results.
■ Infrastructure as a service Developers can build applications from scratch and run them in virtual machines on the cloud servers without having to license tooling that can be costly.
The major advantage of cloud computing is its significantly lower cost relative to the older model, where you would acquire and maintain hardware and software resources. It removes the need for large capital investments in equipment, infrastructure, and software and reduces related operating costs proportionately. It also increases potential mobility because the only thing workers need to do, wherever they are, to access computational resources is connect to the cloud.
The disadvantages of cloud computing are many and include becoming dependent on someone else to control your computational resources. As a consequence, you will fail if they do or if they are unwilling to pitch in to resolve a crisis. Besides other disadvantages, the cloud has serious security and privacy risks, especially if confidential data is not protected adequately. Obviously, there is lots of information available about cloud computing in the professional literature3,4 and on the Internet. (See the end of this chapter for pointers to these resources.)
Your task in this case study is to lead an internal team that has been asked by management to determine whether or not to use cloud computing to provide basic services for your firm and its customers. Customers are residential, industrial, and governmental users of the gas, power, energy, water, and waste-removal services that your firm offers on a fee-for-service basis throughout the municipality.
Organization Figure 6-1 shows an organizational chart of the entire utility company.
The part of the company you are most interested in is the Services organization, which provides a wide range of utility services for a large city and its suburbs. The Information Technology (IT) department of which you are a part provides centralized support for each of these groups from the company’s main offices located in a large city in the southern United States. The department provides infrastructure services for the enterprise as a whole, and automation and operations support services for each of the groups that are part of the Services organization. In addition, IT has a major field site co-located with the services group’s Business and Customer Service (B&CS) organization, which is about 30 miles away. This field site was built as a backup for the main site in case a disaster occurs. Both sites have development facilities and service the firm’s client/server networks. The IT department has 1,200 professionals located at these sites and 300 support personnel. The IT general manager also serves as the chief information officer (CIO) of the company. However, the vice president of operations is responsible for ensuring that the services provided by operations personnel is exemplary at the many sites the company maintains across the country in cities, counties, and government facilities.
The project you are in charge of is primarily tasked with determining whether or not to use cloud computing within the IT department to lower operating costs for infrastructure services like payroll and travel. You have a team of three who are working with you part time to develop a recommendation regarding the use of what management views as a beneficial, cost-cutting technology.
During your team’s kickoff meeting, the IT general manager provides the following added information and direction for the effort:
■ Assess whether moving to cloud computing makes sense for B&CS and Infrastructure Services (IS) because they each maintain various resources that other groups access via the company’s client-server network. The capital assets of each group (equipment, licenses, and other items) are listed in Table 6-1.
■ The number of staff currently required to operate and maintain B&CS and IS facilities and equipment is listed in Table 6-2. These counts address common software used across the company. They do not include the staff used to develop new software applications, who are funded separately by the operational groups.
■ Determine the impact of cloud computing on the IT department’s capital costs of doing business. As part of this analysis, look at what happens to the equipment and facilities that will be disposed of when the department transitions to the use of cloud computing.
■ Make sure that no matter what you do the same quality of service (or better service) is provided to your customers. This is determined by using the current primary measure, which assesses response time relative to a customer’s request for service.
■ Retain any applications that provide the company with clear advantages over the competition, such as the company’s current ability to read meters remotely to tabulate billings.
■ Investigate how to address the security and privacy shortfalls of cloud computing. Make sure customer proprietary information is protected regardless of the cost.
The approach your team decides to take to make your determinations and findings adheres to the practice your firm has in place for such analyses. The steps are outlined as follows:
■ Step 1: Develop a Concept Paper Summarize your concept of operations in a white paper for cloud computing use in B&CS, and summarize its use in IS in a white paper that highlights the capabilities you hope to acquire from the vendors.
■ Step 2: Issue a Public Request for Information Issue a public Request for Information (RFI) asking vendors to comment on your white paper and tell you how they would go about satisfying its requirements.
■ Step 3: Gather Information/Develop Requirements Hold discussions with vendors who provide technically acceptable responses to the list of items you asked them to address in your white paper. These include the evaluation criteria you identified in the paper to be used to make such determinations and findings.
■ Step 4: Prepare/Issue a Request for Proposal (RFP) Issue an RFP to vendors who responded to your RFI for the acquisition of cloud-computing systems and services based on a solicitation that contains your requirements, work statement, and preferred contractual terms and conditions. Make sure that your solicitation does not bias the acquisition by using vendor proprietary information gathered via the RFI process. Otherwise, you might have to deal with a protest from one of the losing vendors.
■ Step 5: Conduct Source Evaluation/Selection Using criteria (responsive to requirements, lowest cost, minimum risk, and other such items) contained within the solicitation, rate and rank the vendor proposals. Make a selection based on the given criteria. Identify major strengths and weaknesses in the winning proposal, and forward it along with your recommendations to those responsible for negotiating a contract.
■ Step 6: Issue a Contract Negotiate with the selected vendor to acquire the products and services using best value and fairness as your overriding principles. Work with the vendor to take advantage of its strengths and compensate for its weaknesses. Remember, the vendor must succeed for you to succeed.
■ Step 7: Monitor the Contractor/Accept Delivery Provide oversight and direction as the contractor works to comply with your requirements for delivery of acceptable products and services. Accept delivery only when the vendor supplies evidence of compliance with your contract requirements.
■ Step 8: Commence Operations Using your concept of operations, transition to the use of the cloud-computing products and services acquired in as disciplined, logical, and risk-free a manner as possible. Apply change-management principles during the transition. Remember to plan in detail because the transition might require you to operate systems in parallel to minimize potential impacts when running a shop 24 hours a day/7 days a week. Be sure to include recurring activities, such as maintaining ongoing communications with stakeholders.
The response to the RFI is overwhelming. You received 42 replies, of which 18 vendors seem to satisfy your requirements. In addition, the telephone has not been idle for the past two days. Most of these phone calls are vendors asking for time to visit and present their wares to you and your team. However, your timeline will not accommodate them all. You need to cut the number to six vendors, at most. You do this by asking all of those responding if they have experience with utility companies similar to yours. Because only three of the six can respond positively, you can reduce the list of promising vendors accordingly.
After discussions with the vendors, you feel that you have the information you need to pull together a briefing to your boss and his staff on how to exploit the use of cloud computing within your utility company. Your briefing will contain the following observations and recommendations relative to the changeover:
■ The move to private cloud computing, where facilities are dedicated to the company, has many benefits as confirmed by government studies5,6 and all of the vendor sales pitches. Costs can be substantially reduced, and the company’s ability to expand and contract its computing resources as needed is greatly enhanced.
■ Given the current business picture and constrained B&CS and IS budgets, private cloud computing seems to represent a viable path forward for the utility.
■ The ability to implement a measured service under a pay-for-use paradigm, which provides services on an on-demand basis across a ubiquitous network, has many advantages.
■ All those interviewed concurred that private cloud computing seems to make the most sense for IS because the services it provides are for the entire organization. Because B&CS is localized, it does not seem to make sense to use private clouds for them on a broader basis.
■ Cost savings will be realized as a function of the substitution of virtualized applications software in the cloud for labor and facilities from current dedicated resources.
■ Cost savings from the cloud can be realized in stages as various applications are replaced by vendor replacements. Transition to private cloud computing will outsource general business applications first and then operations and maintenance later.
■ The team recommends going forward with private clouds for IS but not for B&CS. The next steps in the process will pull together the requirements, develop a solicitation, and issue an RFP for the acquisition of products and services.
■ In preparing for the RFP, the team will identify and seek to retain core services that are fundamental to the way the utility does business and that represent a competitive advantage, such as the ability to read meters remotely and bill clients directly for services, as mentioned earlier.
■ In the RFP, current systems will be kept operational and working in parallel during a transition period of three to five years. Some business process reengineering will be required during the transition to optimize how the new cloud-computing resources are used.
■ In the RFP, as a risk mitigation action, those proposing solutions will be asked how they can address known weaknesses of cloud computing, such as those related to security and privacy.
■ In the RFP, purchase and maintenance options will be included to provide the utility with leverage once the acquisition is completed and the vendor is under contract.
Your boss liked your briefing, nodded his head in agreement many times as he listened, and concurred when you concluded with all of your team’s recommendations. However, the head of IS was infuriated and vocally criticized every one of your charts. Such protests were expected because IS would be taking the brunt of the cuts. When you were asked how much could be saved, you replied, “Based on vendor inputs during the question and answer sessions, they estimated savings between $50 and $60 million of equipment at book value and from 35 to 40 people with a total staff cost between $7 and $8 million. The total reduction based on these numbers is between $57 and $68 million from the current budget of $95 million.”
The head of IS immediately responds to these numbers with a blistering rebuttal. He states that such savings are unrealistic because much of this equipment and the people will have to be retained to run existing facilities in parallel during the three-to-five-year transition period. In addition, at least five new people will have to be hired during this period to perform the business process reengineering tasks, including the staff needed to train users in their proper utilization. He estimates that the conversion costs during the three to five years will add $8 to $10 million to IS’s current operating expenses. These numbers rattle you, the general manager, and the audience. “He is right,” says the general manager. “You need to investigate the costs of transition more fully before I make a go/no-go decision on the acquisition,” he continues.
The head of IS has a grin on his face and looks pleased. He volunteers to have two of his best senior people work with you part time as you develop a response. You politely decline, but the general manager thinks it is a good idea, and you reluctantly accept the offer of help. The general manager schedules an additional meeting two weeks from this one to review the cloud-computing recommendations again.
You should have expected and prepared for the IS response because anticipating and planning for resistance is a fundamental change-management principle. But you did not. Getting cloud computing accepted now will be harder. But it still seems doable.
The team gets together to assess the options with the transition in mind. They identify the following four main transition scenarios to cloud computing that everyone agrees make sense:
■ Option 1: General Application-Only Transition Transition most general applications to a private cloud, retain IS facilities and staff to run general applications, and continue servicing customers on a fee-for-service basis.
■ Option 2: Partial Facilities Transition + Option 1 Perform Option 1, and shut down unneeded facilities within IS. Sell off equipment, and reduce staff proportionately as private cloud services and applications become operational. Upgrade equipment as needed to address reliability issues.
■ Option 3: Transition to Upgraded Facilities + Fuller Set of Applications During the transition to the private cloud, upgrade facilities to provide core processing and backup. Address current equipment reliability issues that are occurring as gear ages and failures increase, thus jeopardizing 24/7 operations. Sell off equipment, and adjust staff proportionately as facilities and cloud services and applications become operational.
■ Option 4: Operate IS As-Is Upgrade IS equipment to address reliability problems, and continue to operate as-is. Perform some streamlining to cut costs and improve service to consumers.
Table 6-3 summarizes the results of the team’s analyses after considerable debate. It identifies the major strengths and weaknesses of each option, along with the estimated costs and projected benefits. The table seems to highlight the overall conclusion that movement to one of the three private cloud-computing options is the right thing to do even though IS remains reluctant to support such a recommendation (that is, the two people assigned to your team neither concurred with this analysis nor agreed to put their names on the results).
Your next meeting with the IT general manager and the head of Infrastructure Services is stormy. It is apparent during the meeting that the head of IS is unhappy with the results. He continuously bombards you during your briefing with nasty remarks, and he accuses you several times of deliberately ignoring his people’s inputs. In addition, he blasts the legitimacy of the numbers and asks for details on how each was derived. You respond with the spreadsheets that provide backup and tell him that his own people reviewed the numbers and found them reasonable.
Your boss finally has no option but to tell everyone to cool down. He states that even though the numbers speak for themselves and seem to present a solid business case for change, he has concerns. His major trepidations, he says, are the risks associated with the transition and the displacement of personnel. He says that equipment has to be changed no matter what option is chosen because it is wearing out and the reliability declines have to be fixed. Based on his remarks, it is not surprising that he accepts Option 2. In response, you and your team take the action to move to the next step in the process by preparing and issuing an RFP. You hope that several of the vendors who replied to your RFI will respond to your RFP with proposals that provide good value for your money.
Your team meets and tries to scope what goes into the RFP besides the requirements and boilerplate text. A member of your team who has been through a large purchase like this before advises you to pay attention to the boilerplate text because this is where the evaluation criteria for selection and the terms and conditions for the purchase go. That’s good advice, you think. So you schedule a meeting between your team and the Purchasing staff.
The meeting with the Purchasing staff goes very well. They had lots of experience and advice about what to put in the solicitation document. Key provisions include rewards for delivering early and penalties for being late. They also provide options to acquire several products and services (additional applications and services, more equipment, training, and other such items) that can be taken after the contract is awarded at a fixed price. Maintenance terms and conditions for the first five years of operations were also spelled out so that you can get the vendor’s immediate and undivided attention when problems occur after delivery.
At the suggestion of the Purchasing department, you send the solicitation out to the prospective suppliers for comment prior to releasing it. You get back a lot of constructive criticism and suggestions. You find that the most controversial clauses are those associated with late delivery and maintenance.
The lessons learned in this industrial case study were many and include, but are not limited to, the following: ■ Even when you think there is a clear choice, resistance to change can pop up from unexpected sources. Therefore, also anticipate resistance and plan to deal with it.
■ Resistance to change comes primarily from those whose power, staff, and budgets are threatened. In this case, such cutbacks are real threats to the Infrastructure Services group.
■ Those who foster change need to anticipate the perceived threats and develop plans to help address them as part of their effort. In this case, figuring out how to find other positions for staff who are no longer needed might have alleviated some of the pain.
■ There might be hidden issues that influence decisions relative to change. In this case, aging equipment and reliability issues did not surface until late in the process when options were being compared. Yet, the issue was one of the major drivers in determining which option was selected.
■ Using competitive market forces to seek the best alternative can be beneficial, especially if you can get an expert review of your solicitation by stimulating the vendors to provide you inputs as to which of your requirements are feasible and which are not.
■ However, relying solely on vendor inputs is dangerous. Because they want to make a sale, they might stretch facts and cloud reality by confusing current capabilities with future capabilities. ■ Using strengths and weaknesses along with costs and benefits permits stronger cases to be made for recommended alternatives.
■ Getting a vendor on contract takes considerable time and effort. It also forces you to solidify your concepts of operations, requirements, and contract terms and conditions.
■ Getting selected vendors to deliver what they promise often takes patience, effort, and due diligence. Many will do a good job. Others may let you down after the contract is issued. To succeed, plan to manage rather than monitor the contract. Otherwise, you probably will get less than what you expect and less than what you are paying for.
■ The challenge will occur after the cloud products and services are delivered and accepted. If you are not one of the vendor’s key accounts, keeping their attention during operations and maintenance might become an issue. That is why I strongly recommend negotiating terms and conditions for any follow-on maintenance contract as part of the original acquisition.
This chapter provides insight into large procurements for Information Technology (IT) products and services. The major issues in this case revolve around addressing resistance to change brought on by a changeover to a new computing paradigm—for example, cloud computing. In this case, such resistance should have been anticipated and dealt with earlier in the process. The team should have gotten IS personnel involved earlier and solicited their inputs and resolved their objections prior to making the recommendation for one of the change options. In the process, they would have learned about and been able to attack the issues of reliability and placement of staff. Instead, they became involved in a war of words that detracted from the goal of the effort, which was determining whether or not cloud computing made sense for this utility company.
This is another chapter where I cut back on materials to save space and maintain a focus. Please understand that cloud computing is controversial and has many issues associated with it that deserve further coverage. For example, the performance of the cloud is overstated—that is, the vendor often promises more performance at a lower cost than it can deliver. As another example, the tools you use may or may not work as advertised on the cloud. Be warned that you need to move carefully to the clouds because they are still in their early-adopter period.