How to identify optimization problems at your company

The following text is an adapted contribution from Feasible, an operations research newsletter written by Borja Menéndez.

Optimization problems are everywhere — across every industry, every level of management, and every business decision-making process.

One of the biggest challenges? These problems grow exponentially.

At first, you throw in a few extra hands to help. But the moment the problem grows even just a little bit, you suddenly need way more people to keep up.

That’s not sustainable. And that’s why it’s so important to spot these problems early — before they spiral out of control. Catch them early, and your business stays on track, doing what it does best.

So today, we’re diving into:

• What exponential growth actually means

• How to recognize optimization problems

• And what steps a business can take to get ahead of them

Ready? Let’s get into it.

📈 Expect the unexpected: exponential growth in optimization problems

An optimization problem with 2,700 decisions that can only be one of two options — like “on” or “off” — might sound easy at first.

It could be something as common as assigning tasks to people — nothing too wild, especially if those tasks follow a sequence rather than being all over the place.

Even something like assigning 20 tasks to just 10 employees already gives you more boolean variables than that.

But here’s the catch: you can’t simply list out all the possible solutions.
Even if you had a computer doing 1 trillion operations per second, it would still take 10³³ years to get through them all 💥

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

Each boolean variable has two possible values: true or false.
With 2,700 of them, the number of combinations is 2 × 2 × 2... (2,700 times) = 2²⁷⁰⁰.

And 2²⁷⁰⁰ is no joke — it’s more than 10⁸⁰⁰, a number so massive it blows past the estimated number of atoms in the universe.

So yeah, listing out every possible solution? Not happening.

The computing power you'd need would outmatch everything we have on Earth — combined.

Now, if you use the right algorithms and mathematical models, you can solve problems like this in just seconds.

But (and here’s the important part): building those algorithms takes time.

So while you're waiting… what happens?
Do you bleed money? Throw more people at the problem? Just hope it gets better?

I’m willing to bet it’s smarter to spot these problems early, before they quietly start sinking your revenue.

🔍 How to identify optimization problems

First things first: you don’t necessarily need to go out hunting for “optimization problems” directly.

What you are looking for are inefficiencies, bottlenecks, and messy decision-making areas — the kinds of issues that slow things down or cost more than they should.
In other words, you're identifying places where complex decisions need to be made better and faster.

Once you’ve done that, you’ll be in a much better spot to decide if Operations Research (OR) is the right tool to solve the problem.
OR is just that — a tool. A powerful one, sure, but don’t use it just because it exists. Use it when it actually helps.

That said, here’s a practical way to start recognizing OR opportunities in your company:

1. Look for Bottlenecks, Inefficiencies, or High Costs

Start by scanning for processes or systems that regularly get stuck or slow things down.
These bottlenecks can be:

• Physical – slow production lines, transportation delays, network congestion, etc.

• Procedural – long decision-making cycles, poor coordination, inefficient task assignment.

Also keep an eye on places where costs are too high — either because of bad planning or lack of visibility. This includes:

• Direct costs like excess inventory, inefficient logistics, or idle machines.

• Indirect costs like unhappy customers due to delays, mistakes, or bad service.

Basically, if you're seeing long wait times, underused resources, overloaded teams, or waste, that's your cue.
It’s probably a sign that optimization — and maybe OR — can make a big difference.

2. Look for complex decision-making environments

Pay attention to situations where decisions aren’t straightforward — where multiple variables, constraints, and stakeholders are in play.
These environments are ideal for advanced decision-support tools like Operations Research.

Here are some classic examples of complex decisions that OR can help with:

• Allocating limited resources — like budget, staff, equipment, or raw materials

• Maximizing or minimizing key metrics — such as cost, profit, time, or output

• Finding the best combinations or configurations — whether it's product assortments, asset mixes, or pricing strategies

• Routing and scheduling — planning vehicle routes, delivery sequences, or production timelines

• Balancing tradeoffs — like improving service levels without blowing up your costs or creating inefficiencies

If a decision feels messy, high-stakes, or time-consuming — there’s a good chance it can be optimized.

3. Consider scalability and impact

Look for areas where small improvements can scale big — where even slight gains could make a noticeable difference for the business or your customers.

Take task planning, for example.
At first, you might think, “Let’s just bring in more people to help.” But as the number of tasks grows, the problem becomes so complex that throwing more hands at it doesn’t help. You don’t get better solutions, and you certainly don’t get them faster.

That’s where Operations Research really shines — in scenarios where automating even part of the process can lead to huge efficiency gains or major boosts in customer satisfaction.

And when we talk about customer satisfaction, we mean everything from faster delivery times to better service experiences and higher product quality — all of which can be optimized with the right models and tools.

👣 Next steps you can take

Alright, so how do you actually do all of this? There are a few solid ways to get started and some key things to keep in mind:

1. Start with your processes

Personally, I like to start with processes. It’s a deep dive ideally done across all departments, but you don’t need to tackle everything at once. Pick one area and work your way through the rest later.

Begin by asking: what decisions have the biggest impact on our results?

Kick things off with in-depth conversations with experienced team members — especially the ones who know the ins and outs of how things really work and also understand the broader business context.

Some great opening questions:

• “Where do you see clear opportunities for improvement?”

• “What are your biggest day-to-day bottlenecks?”

From there, concrete examples will start to surface — like:

• In production: maybe you're dealing with backlogs because machines aren't being used efficiently.

• In warehousing: maybe high picking costs are caused by employees walking too much to retrieve items.

These are exactly the types of pain points that can be mapped, modeled, and optimized.

2. Gather Data and Define the Problem

Start by figuring out what operational data you already have — and what you still need to collect.
Think about things like demand, capacity, costs, travel times, service levels… whatever’s relevant to how your business runs.

If you can’t measure the factors that matter, it’s hard to optimize them — so getting the data right is key.

Next, define what success looks like.
What are the metrics or goals you're trying to improve?
That could mean:

• Cutting costs

• Boosting output

• Improving customer service

• Increasing efficiency

• Or finding the right balance between conflicting priorities

And finally, be clear about your constraints.
List out the real-world limitations you need to work within — things like:

• Budgets

• Machine or staffing capacity

• Labor laws or union rules

• Geographic or service coverage boundaries

These constraints are what shape the edges of your optimization problem — they define what’s possible and what’s not.

3. Get Outside Perspective

While you’re digging into your internal processes, don’t forget to look outward too.
Tap into external expertise — whether it’s through consultants, academic research, or industry case studies.

Consulting firms can give you a view of best practices and emerging trends, while academic or industry literature can show you how others are solving similar problems in creative ways.

Once you’ve gathered all those insights — both internal and external — it’s time to clearly define the core decision problems.
Which challenges could actually be tackled using optimization, simulation, or other OR techniques?

From there, assess:

• What kind of impact each opportunity could have

• The costs and effort involved

• The data you’d need

• And how feasible it is to move forward

Focus first on the highest-value use cases — the ones with the most upside and the clearest path to implementation.

In the end, what matters most is having a deep understanding of your operations, goals, limitations, and data — so you can translate all that into a well-structured, solvable optimization problem.

🏁 Wrapping It Up

Here’s what we covered today:

• Catching optimization problems early is key, especially because they tend to grow exponentially.

• There are at least three solid ways to spot these issues in your business.

• And there are clear steps you can take to start tackling them head-on.

When those complex decision challenges show up, Operations Research is your ally — ready to help you find smarter, faster, and more efficient solutions.

So, what about you?
Have you run into problems like these at your company?
How did you approach them?