📈 #37 How to identify optimization problems at your company

So you can save money on solving them.

I know it, you know it too, but there’s a lot of people out there that don’t know it yet.

Optimization problems are everywhere. In any industry, in any level of management, in any decision-making process we find in a business.

One of the main issues is the exponential growth of these problems.

You sometimes throw more hands to solve the problem and as soon as it grows just a little super tiny bit more, you need more and more hands.

This is not sustainable for any business, and it’s critical for them to spot these problems even before they show their full potential. That way, the business is safe and is able to continue doing its thing.

So today in Feasible we are going to talk about:

• What’s exponential growth

• How to identify optimization problems

• The next steps a business can take to address them

Ready for today? Then… Let’s go for it!

📈 Expect the unexpected: exponential growth in optimization problems

An optimization problem with just 2700 boolean variables seems easy to solve at first sight.

This can easily be an assignment problem of tasks to people, for example. And not specially a big one when there’s a sequence of tasks instead of isolated tasks.

Having just 10 employees to assign 20 tasks is a problem with even more boolean variables than I said before.

But let me tell you one thing: you cannot enumerate all possible solutions. A computer that performs 1 trillion operations/second will need 10³³ years to complete 💥

That’s why you should never understimate the power of exponential growth.

• Each boolean variable can take on 2 values: true or false

• With 2700 variables, there are 2 choices for each variable

• So the total number of possible combinations is 2 x 2 x 2.... (2700 times) = 2²⁷⁰⁰

• 2²⁷⁰⁰ is an incredibly huge number: it's over 10⁸⁰⁰

It's impossible to list every solution for a problem of that magnitude.

The computational power needed would surpass everything on Earth combined.

If you use algorithms and mathematical models to solve that kind of problems, it will take you just some seconds.

But (and here’s the thing).