“Optimization” is a sexy buzzword that gets thrown around a lot in the business world. Phrases like “Let’s optimize our advertising spending” or “start a project to optimize our patient flow” get used a lot without a real understanding what the word or method means.
As an industrial engineer, I studied optimization in my undergraduate studies at Northwestern University. Optimization is pretty rigorous from a mathematical standpoint, and we used computer software to build optimization models. When you’re optimizing an equation, you generally have a single goal or objective, such as cost or travel distance for trucks that you’re trying to maximize or minimize, given a number of constraints. Things quickly get complicated if you’re trying to optimize multiple variables, especially when there are tradeoffs to be considered.
Optimization and “linear programming” is a pretty complex field. Yeah, you can build simple optimization functions in Microsoft Excel, but it’s generally something that’s done by specialists.
If a business were to really “optimize” their ad spending, they would need to make a number of assumptions to build the model, such as how many people will buy products as the result of TV ads versus Facebook ads (and some of those assumptions might be wrong). If a hospital team wanted to “optimize” patient flow, they’d be trying to solve for a feasible number that represents the shortest possible patient length of stay.
Anyway, I’m a bit rusty on the technical details after 20 years. But, I’m rusty because people generally don’t use true optimization methods in the real world. They aren’t actually finding the optimal answer… they’re just setting stretch goals, and they’re improving.
Instead of looking for the “right” answer, organizations use continuous improvement or optimization methodologies to experiment, constantly searching for ways to get a little better. We can adjust our ad spending budgets and see the results of those changes. We can run Kaizen Events or other projects that will reduce length of stay for patients… little by little, getting better every day, week, month, and year.If you hear someone say “let’s optimize that,” get to work improving! Instead of working toward some theoretical ideal “optimal” number, schedule some Kaizen Events, teach some PDSA problem solving, and measure your results. Trust me, you’ll get further… even if you disappoint an industrial engineer in the process.