Growth Requires Solving Both Scalable and Non-scalable Problems
Note: We’ve just completed Part 2: The Right People for the Right Problems—exploring why managers and entrepreneurs think differently, why optimization alone leads to stagnation, and why markets require local adaptation rather than universal solutions. Now we’re starting Part 3: New Tools, New Roles. These chapters show how to combine different approaches for sustainable growth. Today’s chapter: how to tell which problems can be solved with scalable systems and which require non-scalable, context-specific solutions—and why confusing the two kills both efficiency and innovation.
One Takeaway: Businesses face both scalable and non-scalable problems. Leaders must recognize when to use structured models and when to rely on judgment.
Real vs Ideal
The real world is not the ideal world. If it were, companies would have perfect knowledge of customer preferences, competitive landscapes, and future demand. In that world, math could resolve every decision. But businesses don’t operate in equations. They operate in uncertainty.
Our choice is not between perfect or imperfect; it’s between better or worse. Our choices can either lead to growth or stagnation. Because of this, we must choose wisely.
Why This Matters
Ideal assumptions like stable markets, complete data, and predictable conditions often form the foundation of many decision-making tools. When used correctly, these models are limited, but valuable. When misused, they lead to rigid thinking, missed opportunities, and false precision.
Models that can be helpful can become harmful when used in the wrong situations. Businesses don’t just need models when solving problems, they also need judgment. Leaders must tell the difference between scalable and non-scalable problems to determine the right approach.
Scalable vs. Non-Scalable Problems
Scalable Problems: Are challenges that can be solved through structured, repeatable solutions. They benefit from efficiency, automation, and data-driven decision-making.
Non-Scalable Problems: Are challenges that require customized approaches. They come from unique market conditions, shifting consumer behaviors, or new opportunities. They don’t always fit into models, making experience, gut instinct, and experimentation essential.
Quick Check: Which Type Are You Facing?
Ask these questions to identify whether you’re dealing with scalable or non-scalable challenges:
Signs you’re facing a scalable problem:
Past data reliably predicts outcomes that you can observe and track
Similar solutions work across different situations
Making processes standard improves results
Success factors are clearly measurable
Signs you’re facing a non-scalable problem:
Each situation has unique characteristics that affect outcomes
Local knowledge creates advantages that data misses
Best practices from elsewhere don’t work in your situation
Testing reveals more than analysis
Real-World Examples
At Uber and Lyft, dynamic pricing and route optimization were clear examples of scalable solutions. These tools ensured consistency across markets. However, when launching in new cities, local team insights often proved more valuable than centralized data models.
Market entry required judgment, adaptation, and non-scalable work. These are all things formulas can’t predict. But many decision makers and business leaders prefer focusing on collecting more data and calculating more models because they seem more sophisticated. This desire for sophistication can lead to over-reliance on data and under-reliance on thinking or action.
As Charlie Munger famously warned: “People calculate too much and think too little.”
Making the Non-Scalable Scalable
Sometimes, non-scalable solutions evolve into scalable ones. Many successful business models start as unscalable, high-touch experiments before being refined into repeatable processes.
Uber Eats began as a small experiment in Santa Monica called uberFRESH. The design didn’t aim for immediate scaling. It tested demand and logistics in a controlled environment. Only after proving it worked did Uber refine it into a global food delivery network.
Other industries follow similar patterns:
Retail: Pop-up stores test demand before scaling to permanent locations
Hotels: Boutique hotels experiment with unique experiences before chains adopt successful concepts
Tech: Startups launch with hands-on customer service before automation enables broader expansion
Data Needs Judgment
Models require interpretation. As economists Matt Mitchell and Peter Boettke note, “Measurement without theory is impossible... data require interpretation.”1
Raw data doesn’t give answers by itself. It requires judgment to be useful. Companies that blindly follow analytics without understanding market context risk making poor decisions.
Businesses leaders must question:
What data is necessary?
What data holds you back?
Where and when is judgment more valuable than measurement?
The Right Tool for the Right Problem
Scalable problems benefit from structured frameworks and optimization approaches.
Non-scalable challenges require flexibility and experimentation.
Business leaders must recognize models as tools, not solutions. Over-reliance on any single approach, whether that be data-driven optimization or pure gut instinct, limits growth.
Examples across functions:
Logistics: Computers optimize delivery routes, but unexpected problems require human intervention
Product development: A/B testing refines features, but breakthrough innovations come from judgment and risk-taking
Customer experience: Chatbots handle routine requests, but service recovery depends on human problem-solving
Beyond “If You Can’t Measure It”
People often misunderstand the old saying, “If you can’t measure it, you can’t manage it.” Just because something can’t be easily counted doesn’t mean it isn’t valuable. Some things aren’t meant to be managed, they’re meant to be created.
As we’ve discussed before, creating something new vs managing something that already exists are two very different decision making environments.
The Bottom Line
Managers and entrepreneurs use different tools to solve different problems.
Managers value structured models, efficiency, and optimization. These are perfect for scalable solutions.
Entrepreneurs embrace uncertainty, rely on judgment, and create value where models don’t yet apply. These are essential for non-scalable challenges.
The key is knowing when to use each approach. Combining these approaches—using models where they work well (not where they work perfectly) and judgment where they don’t—helps businesses both optimize and innovate.
Now that you understand the difference between scalable and non-scalable problems, the question becomes: how do you actually combine these different approaches in practice? How do you build an organization that excels at both structured thinking and creative problem-solving?
Mitchell, Matt, and Peter Boettke. Applied mainline economics: Bridging the gap between theory and public policy. No. 06895. George Mason University, Mercatus Center, 2017.