Reference no: EM133370975

A demand forecasting system Case Study

Pat, your manager, has returned from an MIT course, excited at having learned about new methods of demand forecasting for your organization. Your manager calls a meeting about the new project. "This demand forecasting system, which we're buying from a vendor but also modifying for our organization, is critical to our future. We've anticipated demand changes very poorly in the past, and we have not had sufficient capacity in the right places at the right time to meet our customers' needs. I know that our organization hasn't supported logistics and transportation system efforts very well in the past, so this time we'll give you everything you need. You'll have good office space, good servers and software support, and you won't have to attend most of the regular staff meetings." "That sounds excellent," you say. "My staff, which is all recent MIT SM grads, is really enthusiastic. They don't have as much experience as I'd like, but their energy and smarts will more than make up for it. I'll get commitments from them to work really hard in exchange for allowing them to be really creative, so that we complete the project on time." "I've budgeted this as an 8 month project, and I've promised the Board of Directors that we'll be done on time," Pat said. "This is a very high visibility effort." Alex, from QA, says, "If we're going to be using the system at the 8 month point, we need to have the software frozen at the 6 month mark, so we can write the test scripts. Since your system will interface with our current customer measurement software and our geographic information system (GIS), we'll need time to test the interfaces." Chris, from training, also chimes in: "We'll need the system frozen at the 6 month point also, to prepare training materials for the staff." You agree to the 6 month freeze. As your staff of 4 analysts begins the project, they're very energized by the challenge and the freedom of the effort. They split up the project into pieces: Joe defined the interfaces with the other systems, and Karen designed the database and system interface components to achieve this. Kate configured the statistical modules of the vendorsupplied package, and specified some additional analytics that Jim would writ3 as extensions. Karen liked to design user interfaces, so she took on that task. You managed the project using a spreadsheet and simple project management software (Microsoft Project). The staff took one month to be trained in the new software and to talk to the groups who did the current customer measurement and GIS systems. They discovered that most, but not all, the data needed to improve the demand forecasts was available, and that some existing data was coded incorrectly and grouped in ways that didn't work well. They negotiated changes with the customer measurement and GIS system groups, who promised the changes would be complete in two months (at the 3 month point in the project). The interfaces had to be redesigned when the other groups completed their work. Because they had to make the changes quickly, the other groups tweaked the existing interfaces and queries to produce what was needed, even though in some cases much more data was sent than needed, and in other cases, the data had to be reorganized within the demand forecasting system to be usable. This took all of month 4, and a few compromises had to be made. The statistics modules then had to be redesigned because the compromises affected the models; this took month 5, and then month 6 to take care of some nagging problems. There were still a few spots that didn't seem to work properly, but the staff could make them work with care. The user interface then had to be redesigned, and this took through month 7. Because the user interface wasn't ready, neither QA nor training could start their work. You said, "Well, you really shouldn't need two months for just QA and training. It shouldn't take 25% of the effort just to do that; we've been busting our butts to build all this. You can do it in one month." You go to your boss, and say that you really need her support in this, and that she'd promised you a lot of support for this project. She reluctantly agrees, and tells QA and training they have to be done in one month. She tells the Board the project is still on schedule. Your team finally finished the first version of the customized software at the 8 month point and hands it over to QA and training. They can't get the system to run properly. Your staff sits with them and points out all the parameters that must be set in just one way, and the exact sequence in which all the analyses must be run. It's very complex, and QA and training both say that the users can't deal with it. Also, any small error results in system crashes or crazy results. An impasse results; you all meet with Pat the next week, during month 9. Pat asks for a demo of the system, which you haven't been able to provide until then. After the demo, she says that she can't understand how the system works and she couldn't use it either. She clears her schedule, and sits with the five of you, plus QA and training, for the remaining four days that week, simplifying the system, throwing out many sophisticated models that don't quite work, streamlining the system interfaces (making them just one way into the forecasting system, with nothing going back to the other systems, as originally intended), and eliminating half the user interface Web pages. She's mad. She tells your team to work 12 hours a day, 6 days a week, and to finish the simpler system by the end of month 9. Your staff grumbles mightily but is embarrassed and does it. You finish in the middle of month 10. QA and training take the system, and find many bugs and inconsistencies. They also work overtime for the next 6 weeks and 2 you finally deliver a simple system, not much different than the 'out of the box' configuration, at month 12. You get a 1% raise for the year, and you feel very underappreciated. But when you interview for other jobs, no one is that excited to hire you.

Questions: 

a. List as many errors that were made by this team as you can. 

b. What did the team do right? 

c. What project management method was used? Was it appropriate? 

d. What should they have done to succeed