Why Growing Companies Need More Than Specialists
The Economic Logic Behind Operators, Refiners, and Creators
One Takeaway: We learned in Growth Isn’t One Sided that organizations need operators, refiners, and creators to succeed. But why can’t you just hire good specialists and let them figure it out? The answer lies in economic principles about specialization and coordination that explain why these distinct types of work are needed.
The Big Picture: In Growth Isn’t One Sided, we established that companies tend to naturally develop three distinct types of work: operators who maintain stability, refiners who optimize existing systems, and creators who discover new opportunities. We saw how treating all three the same way can kill growth, while designing different systems for each can enable sustained success.
But this raises deeper questions: Why do these three functions emerge so predictably across different organizations? Why do talented people often struggle to just adapt their approach or what they focus on based on what the situation requires? And most importantly, how can you diagnose whether your organization is properly supporting all three functions or accidentally undermining them?
The answers come from economic principles (surprise!) about how specialization creates value and why coordination becomes more challenging, and more essential, as organizations grow. Understanding the underlying economic logic helps you adapt the frameworks to your specific situation rather than just following templates.
Why Specialization Creates Value (And Some Problems)
The economic logic behind operators, refiners, and creators traces back to an important insight from Adam Smith about how wealth gets created. In 1776, Smith observed that ten workers could produce tens of thousands of pins per day through each specializing, versus maybe a couple hundred pins if each worker made complete pins individually.
This wasn’t simply about assembly lines. This insight revealed something fundamental about how specialization creates economic value that wouldn’t exist otherwise. But Smith’s insight goes beyond physical tasks to what we could think of as “thinking specialization.” Just as dividing physical work increases manufacturing productivity, dividing thinking work can increase organizational effectiveness.
The three functions from Growth Isn’t One Sided represent “thinking” specialization in action:
Operators specialize in maintaining effectiveness. They develop expertise in keeping existing systems running predictably and reliably. This enables them to spot problems before they become failures, search for ways to make routine processes consistent, and handle the complexity that would overwhelm people trying to think about everything at once.
Refiners specialize in efficiency improvements. They develop expertise in analyzing existing systems and finding ways to make them work better. This enables them to identify inefficiencies that others miss, design experiments that reveal optimization opportunities, and balance multiple variables to improve overall performance.
Creators specialize in discovering possibilities. They develop expertise in recognizing opportunities and experimenting with approaches that don’t fit existing systems. This enables them to notice market changes before they show up in data, identify customer needs that aren’t being served, and develop solutions that didn’t previously exist.
Coordination is difficult because operators, refiners, and creators literally think differently about problems.
Operators think in terms of immediate execution, refiners think in systems and processes, creators think in possibilities and uncertainties.
Bridging these different ways of thinking, translating operator concerns into refiner language into creator opportunities, requires a rare capability we’ll explore later when we dive more into Harmonizers.
But This Isn’t Just Division of Labor—It’s Division of Problem Spaces
Here’s where thinking specialization differs fundamentally from manufacturing specialization. In Smith’s pin factory, workers perform sequential tasks on the same product. One person draws the wire, another straightens it, another cuts it. Modern manufacturing has learned that extreme task specialization can create fragility. When one step breaks, the whole system can screech to a stop.
Why doesn’t the same fragility affect operators, refiners, and creators?
Because they’re not doing sequential tasks on the same problems. They’re addressing fundamentally different types of problems that exist simultaneously in every organization:
Operators address effectiveness problems. They keep existing systems stable and reliable
Refiners address optimization problems. They make existing systems work better within known constraints
Creators address discovery problems. They identify opportunities and solutions that don’t yet exist
These aren’t steps in a single process. They’re distinct economic challenges that require different modes of thinking from one another. You likely can’t “discover new opportunities” and “maintain system stability” using the same mental framework, any more than you can simultaneously accelerate and brake in your car.
The specialization works because each function tackles a different category of business challenge, not because they’re dividing up a single task. This is why thinking specialization creates value rather than fragility. The functions are genuinely independent problem spaces that happen to need coordination to work together.
The coordination challenge: Just as Smith’s pin makers needed someone to coordinate their specialized activities, organizations need ways to coordinate between these different types of thinking without destroying the benefits that specialization provides.
Why Just Hiring Good People and Let Them Figure It Out Usually Doesn’t Work
Consider a 150-person software company growing 40% annually. Their customer success team keeps the platform stable with 99.7% uptime. Their product team improves conversion rates through A/B tests.
Revenue is growing, investors are happy, and everyone seems aligned.
Then a major competitor launches AI-powered features that fundamentally change what customers expect. The company needs to respond, but their organizational systems are structured around maintaining and optimizing the current product. Nobody’s metrics reward “explore what AI capabilities we should build over the next 18 months.”
The customer success team (the operators) can’t experiment with undefined features because their job is stability. The product analytics team (the refiners) can optimize AI features once they exist, but creating the initial strategy doesn’t fit their framework. And unfortunately, the few people thinking creatively about AI possibilities don’t have budgets, decision authority, or success metrics that reward their time considering these opportunities.
The opportunity is obvious. But the structure of the organization makes pursuing it irrational for everyone involved.
This happens because the reason you need distinct support systems for operators, refiners, and creators isn’t about personality types or individual capabilities. It’s about the structure of different types of problems they each face.
When Problems Have Different Information Structures
In 1921, economist Frank Knight made a crucial distinction in his work Risk, Uncertainty, and Profit that explains why some business challenges require calculation while others require judgment. He separated risk from uncertainty.
Risk means you don’t know what will happen, but you know what could happen and roughly how likely each outcome is. You can analyze historical patterns, run experiments, and calculate expected values. Think of optimizing pricing strategies or improving conversion funnels. With those systems you have enough reliable information to use straightforward analytical approaches. A good example of this is a standard deck of 52 cards. You know how many Kings there are, so you can calculate the probability you will get a King at any given time.
Uncertainty means you don’t even know what outcomes are possible, let alone their probabilities. You’re operating in genuinely new territory where historical patterns provide limited guidance. Think of evaluating whether to enter a new market or how AI will reshape your industry. Analysis can help to a degree, but judgment ultimately must drive your decisions. A good example here is you have a deck of cards and you don’t know how many are in the deck and you don’t know if there are even any Kings in it and the deck may change from time to time.
This isn’t a subjective distinction or an excuse to avoid rigorous thinking. It’s an objective feature of different types of problems businesses face.
Operators work in relatively low-uncertainty environments where stable relationships between inputs and outputs mean historical patterns reliably predict future performance. When you’re keeping systems running, you want repeatable processes based on what’s worked before.
Refiners work in moderate-uncertainty environments where variables are measurable and respond to analysis, but relationships aren’t perfectly stable. When you’re optimizing conversion rates, you need experimentation, but you’re still working within established systems where you can measure results.
Creators work in high-uncertainty environments where existing data provides limited guidance about what might be successful. When you’re exploring AI strategy or identifying unmet customer needs, you’re discovering what’s possible, not optimizing what exists.
Trying to use the same management approach for all three violates basic economic principles about matching your decision-making method to your information environment. It’s like trying to use only a screwdriver to build a house. You get poor results not because the builders aren’t capable, but because screwdrivers aren’t good at cutting 2x4s or sanding plywood. The approach doesn’t fit the problem structure.
When companies try to apply analytical, optimization-focused management universally, they’re essentially treating all problems as risk problems when many are actually uncertainty problems. The tools that work brilliantly for risk (data analysis, A/B testing, ROI calculations) can easily fail or mislead when applied to genuine uncertainty.
Why Traditional Management Fails All Three
Most companies default to what we might call “refiner logic.” They tend to focus on optimization and improvement and use it everywhere. Why? Because it’s the approach that’s easiest to measure, justify to stakeholders, and scale through established management practices. MBA programs teach it. Consulting firms sell it. Investors understand it.
But economic logic shows why what seems like a no-brainer default choice can create systematic failures across all three functions. When you apply optimization frameworks universally, you create what could be seen as “internal market failures” (for lack of a better word) where the wrong incentives guide how people spend their time and organizational resources (other economists may take issue with this usage, but I think it’s helpful even if not 100% correct) .
Operators get frustrated because optimization approaches can jeopardize effectiveness. Continuous new initiatives can disrupt the consistent processes that reliable operations depend on. Performance reviews that emphasize innovation and change can penalize the steady execution that keeps businesses functioning. Short-term optimization can sometimes undermine long-term stability.
Refiners get overwhelmed when they’re asked to both optimize existing systems and discover completely new approaches. Systematic improvement requires different skills, information, and time periods than innovation. Asking refiners to also function as creators often means they do neither well.
Creators get killed by optimization systems because discovery work doesn’t fit improvement frameworks. Innovation requires tolerance for failure, longer time horizons, and different success metrics than operational work. Traditional measurement systems designed for efficiency often punish exactly the experimentation that drives long-term growth.
The Hidden Economic Costs of Misalignment
When organizations fail to recognize operators, refiners, and creators as distinct functions that require different support systems, they create three types of costs that compound over time.
Opportunity Costs Accumulate
When creators are forced to focus on operational metrics instead of discovery work, organizations optimize for yesterday’s opportunities while missing tomorrow’s possibilities.
Remember the retail chain example from Growth Isn’t One Sided. In that example only the creators were addressing the demographic shifts that represented the real business challenge. When measurement systems prevent creators from doing creator work, you’re not just losing their productivity. You’re losing their unique contribution.
This is where the economic principle of comparative advantage becomes critical. Even if a creator can do operational work competently, and many can, having them do so represents massive opportunity cost. Similarly, having operators try to do creator work isn’t just inefficient; it’s economically destructive to the organization’s total capability.
The economic logic is straightforward: Organizations maximize value by having each function work on problems where they have comparative advantage, not just absolute advantage.
Think about it this way. Your best creator might be a very good operator. They could maintain systems adequately if needed. But every hour they spend on operational work is an hour not spent discovering opportunities that only they’re positioned to identify. That’s not just an efficiency loss; it’s a strategic loss of competitive advantage.
Similarly, your best operators might be capable of creative thinking. They could probably come up with innovative ideas if pressed. But forcing operators into creator work undermines both functions. Operators lose focus on the stability work that’s their comparative advantage, while creator work suffers from being done by people whose thinking specialization makes them better suited for different challenges.
This means even cross-functional people who can legitimately operate in multiple modes should be assigned to functions based on where they create most value relative to alternatives, not just where they’re capable of contributing.
Coordination Costs Multiply
The “Bill problems” that fall between departments often persist because coordination systems designed for operational work don’t accommodate the cross-functional, experimental approaches that many business challenges actually require (economists call these “transaction costs”). We’ll explore this more deeply when we discuss Harmonizers, but for now, it’s best to understand that misalignment doesn’t just hurt individual functions. It makes the functions unable to work together effectively.
You can lose both the forest and the trees at the same time.
Innovation Gets Systematically Underinvested
Projects that could create significant long-term value get killed because they don’t generate immediate measurable returns, while incremental improvements get over-funded relative to their strategic importance. The result is that organizations become excellent at optimizing their way to irrelevance while competitors who maintain a focus on new product or market discovery adapt to changing markets.
Economic Diagnosis: Is Your Organization Supporting All Three Functions?
Understanding the economic logic helps you diagnose whether your organization is accidentally undermining the functions it needs. Rather than waiting for outcome failures (quality problems, competitive losses, missed opportunities), look for the behavioral signals that show misalignment is happening:
Early warning signals you’re under-supporting operators:
Operators requesting “exceptions” to improvement initiatives to maintain stability
Reliability metrics trending down while efficiency metrics trend up
Operators spending increasing time in meetings defending “why we do it this way”
People saying “we aren’t focusing enough on execution”
Early warning signals you’re under-supporting refiners:
Refiners unable to get time/resources for analysis because “we need to ship”
Known inefficiencies persist for multiple quarters without attention
Refiners’ improvement proposals dying in “prioritization” meetings
People saying “we are clearly wasting resources here”
Early warning signals you’re under-supporting creators:
Creators feeling forced to spend time reframing discovery work to fit optimization metrics
Innovation initiatives consistently producing incremental rather than new solutions
Creator-type people leaving to join earlier-stage companies
People saying “we need to think outside the box”
Early warning signals your coordination between functions is failing:
Operators and creators constantly in conflict about priorities
Refiners unable to get information they need from other teams
Good ideas that die in implementation
People saying “we need better communication”
What To Do About It
If you’re under-supporting operators: Create “protected” time where process improvements aren’t interrupted by other initiatives. Let operators say no to “efficiency improvements” that they feel will sacrifice their job to provide reliable results. Recognize that keeping the lights on and running the ship smoothly itself is a form of value creation, not just simply the absence of having any problems.
If you’re under-supporting refiners: Give them dedicated time for analysis and experimentation within your existing business. Success metrics should reward improvements to core processes, not just new feature launches. Ensure they have access to the data and experimental capabilities needed for optimization work.
If you’re under-supporting creators: Separate discovery budgets from operational budgets. Evaluate creator work on learning and option creation, not immediate ROI. Give them permission to pursue opportunities that don’t fit departmental OKRs. Remember that being smart about trying new things and failing creates valuable learning.
If coordination is failing: This is where the Harmonizer concept from Growth Isn’t One Sided becomes essential. You need people explicitly focused on connecting these different types of work toward shared outcomes rather than letting each function optimize independently.
And above all, throughout all this, document everything and build a knowledge base for your current and future teams to draw from. The work you do isn’t one-and-done. The more you document and set a standard for learnings the better your future efforts will likely be. This is the true value in working in teams within growing organizations.
Different Functions Need Different Support
The solution isn’t choosing between these functions or hoping people will naturally balance them. It’s designing organizational systems that recognize the different economic logic of each type of work.
Operators need systems that reduce coordination costs and support consistent execution. Clear processes, reliable information flows, and predictable workflows enable the stability that operational effectiveness requires. Success metrics should focus on reliability, quality, and efficiency within established parameters.
Refiners need systems that enable analysis and improvement within existing frameworks. Access to performance data, experimental capabilities, and improvement targets enable the optimization work that drives effective and efficient operations. Success metrics should focus on measurable improvements to established systems and processes.
Creators need systems that support exploration and experimentation under uncertainty. Discovery budgets, longer evaluation timeframes, and tolerance for failure, within guardrails, enable the entrepreneurial work that drives adaptation and long-term growth. Success metrics should focus on learning, opportunity identification, and option creation rather than immediate measurable returns.
Most importantly, you need coordination systems that help these different functions work together without undermining each other. This is where the Harmonizer concept becomes essential. Harmonizers are not just a role. They are a function that solves coordination challenges between different types of specialized thinking.
Adapting the Framework to Your Context
Understanding the economic logic enables you to adapt the operators, refiners, creators framework to your specific situation rather than just implementing it generically:
In stable industries, you might need more operator capability relative to creator capability, but you still need both. The economic principles remain the same even when the proportions change.
In dynamic industries, you might need more creator capability relative to operator capability, but reliable operations still matters for customer experience and competitive positioning.
In different organizational contexts, the same person might serve different functions at different times, but the functions themselves remain distinct and require different support systems. Remember: assign people based on their comparative advantage, not just their capability.
At different organizational scales, you might organize these functions differently, but the underlying economic logic about specialization and coordination remains constant.
Where We’re Going Next
Understanding why these three functions exist and need different support systems is the foundation. But how do you actually decide which problems require operational thinking, refinement thinking, or creative thinking in practice?
That’s where some of my favorite economists provide crucial insights. Some problems have enough reliable information that calculation is a viable (but imperfect) approach, while others have so much genuine uncertainty that judgment is the only economically sound method. Next week, we’ll explore some thoughts on when to use each, and why getting this wrong can destroy value even when you have the right people doing the right work.
The Bottom Line
The operators, refiners, and creators framework works because it reflects fundamental economic principles about how specialization creates value and how coordination enables specialized functions to work together effectively.
Understanding this economic foundation helps you use these concepts more effectively because you can adapt the approach to your specific context while preserving the underlying logic that makes it successful.
The competitive advantage comes not just from having all three functions, but from designing org systems that enable these thinking specializations while preventing coordination failures that can undermine effectiveness. This all comes from simply applying the principle of comparative advantage internally in your company. You must put your operators, refiners, and creators to work on the problems where they create the most value relative to the alternatives available to them.