After lunch, we heard from Jon Walkwitz from one of the largest energy companies. He's since completed several network optimization projects using ILOG LogicNet Plus and ILOG Inventory Analyst. He leads the team's monthly sales and operations planning process.
Jon talked about several optimization projects that he'd worked on, including customer aggregation and product aggregation projects, with a goal of minimizing transportation costs out to the company's customers while keeping production costs neutral. He went into detail on the types of data that go into the modeling process and the level of complexity involved in the process. His point was that the level of detail available through using an optimization is not available when using a spreadsheet approach.
He noted barriers to achieving savings, such as "disaggregating and identifying actual direction from noise," "capital required to achieve savings" and "cost of doing business." He repeated how difficult the planning process can be, and you have to recognize that it will take a significant effort to collect, analyze and feed the model. You need a dedicated team to maximize efficiency, and you have to manage shareholder expectations and perceptions so that your internal customers don't come into the process believing that you can plug numbers into a black box, wave your magic optimization wand and produce millions in savings.
Jon also discussed the company's Inventory Analyst project and some of the barriers to achieving results in this sort of project, including conceptual difficulty understanding the project, which makes it that much more difficult to win stakeholder buy-in. He also noted that the models can be difficult to understand, and the models can help answer questions but don't necessarily inspire epiphanies – there's not necessarily an "easy button" that magically produces results. So you need to ensure that you hire the right team to drive the models and explain the results, including, if necessary, hiring a consulting team at least initially until you get comfortable with the process, in part to help you understand and explain the results.