Carrier Corporation: Redesigning the HVAC Giant
This is post 22 of 23 in a series on “Systems Thinking: Managing Chaos and Complexity” by Jamshid Gharajedaghi (ISBN 978-0-7506-7973-2).
Previous: Commonwealth Energy - Business Units and Services
How a Phone Call Almost Didn’t Happen
Gharajedaghi opens this chapter with a great story. He was invited to present to Karl Krapek, Carrier Corporation’s CEO. The deal was simple. He’d be paid for a full day, but if Krapek picked up the phone to make a call, the presentation was over.
Fifteen minutes in, Krapek reached for the phone. Gharajedaghi started packing up. But Krapek wasn’t ending the meeting. He was calling colleagues to come listen. That one phone call turned into a full redesign project for one of the world’s biggest HVAC companies.
The challenge was massive. Carrier had over 90 autonomous operations scattered across the globe. They needed to become one system, not just a collection of parts. Their core technology needed updating. Their distributor network was aging and resistant to change. And competitors were getting faster, cheaper, and smarter.
The Rules Were Changing
Gharajedaghi lays out the shifts happening in industry at the time. Mass production based on economy of scale was giving way to flexible production with low break-even points. Companies couldn’t just manage variables one at a time anymore. Everything was connected. Knowledge wasn’t enough on its own. You had to actually put that knowledge to work.
Here’s one that still hits hard today: the shift from target pricing to target costing. Instead of setting your price and hoping the math works out, you figure out what the market will pay and then design your costs to fit inside that number. The cost becomes what you control, not the price.
And labor was changing too. The old view treated workers as variable costs. Cheap labor was a competitive advantage. The new view treated workers as fixed assets. Knowledgeable workers were the real advantage, not disposable ones.
HVAC Was Getting Squeezed
The HVAC industry had its own set of problems on top of all this. Strong component suppliers were emerging. These companies had the scale and focus to control key parts like compressors. They could set prices and even threaten to become competitors themselves by selling end products directly.
Then you had the “no frills” competitors offering rock-bottom prices with lean operations. And Japanese companies were combining great operations with high-tech products. Their playbook was clear. Start at the lower end of the market, offer easy-to-use packages that need almost no maintenance, sell directly to retailers (skip the distributors entirely), and then move up to bigger products.
Electronic controls were becoming a must-have. Environmental concerns were growing. Markets were maturing in some places and emerging in others. New tech like duct-free splits was reshaping the entire value chain.
What Carrier Decided to Stand For
The design team defined some clear core values. They weren’t just going to make equipment. They were going to be designers, producers, and distributors of climate control systems. Everyone in the chain, from distributors to end users, was a customer.
They committed to guaranteeing satisfaction all the way to the end user. If you bought a Carrier product, even through a private label, Carrier stood behind it. No excuses. They also made a bold move on job security. Nobody would lose their job because of plant closings, restructuring, or market downturns. That takes guts.
The team also drew a line on competence. They valued excellence through knowledge. They promoted, and I’m quoting here, “intolerance of incompetence.” It’s a strong phrase. In a company with 90+ operations, it sends a clear message about expectations.
Core Technology Priorities
The team ranked their core components in order of importance: compressors, electronic controls, heat transfer devices, enclosures and fans, motors, diesel engines, and air cleaning devices. They categorized their tech needs into three buckets.
Some things they had to know and control in-house, like compressors and electronic controls. Others needed knowledge plus strategic alliances, like heat transfer and motors. And some just needed enough knowledge to influence outside developers, like refrigerants and building software.
One point stands out. They wanted to replace their “build and test” culture with a “scientific predictive modeling” culture. Instead of making something and seeing if it works, model it first. Train engineers to think across disciplines, not just within their specialty.
The Three-Dimensional Architecture
This is where the systems thinking really kicks in. Carrier’s redesign used a multidimensional architecture built on three dimensions: markets, products, and technology. The whole point was to avoid what Gharajedaghi calls suboptimization. That’s when you focus so hard on one thing that you hurt everything else.
The architecture needed centralization and decentralization at the same time. Integration and differentiation. That sounds impossible, but the structure they designed actually makes it work.
Markets were divided into four semi-autonomous areas, each managing regional units. A region was defined by climate, construction practices, economic development, and physical access. Regional units focused purely on sales, distribution, and service. They didn’t manufacture. Their job was to understand end users and feed that information back to the product teams.
Area Units sat above regional units. Each area had three segments: residential, commercial, and industrial/institutional. Area units had no fixed assets on purpose. If you own factories, you think about factories. If you don’t, you think about customers. They constantly evaluated “make or buy” decisions, keeping competitive pressure on both internal manufacturers and outside suppliers.
Output Units handled design and production of global products. These were profit centers with freedom to sell and source externally. They used a global platform concept. Think of it like modular design. You standardize the core chassis and components, then allow variations for local requirements. One design, many markets.
Components, Technology, and Support
The Components Group managed critical parts like compressors and electronic controls. Each component business had to survive on its own. They could sell to outside manufacturers, and Carrier’s product managers could buy from outside suppliers. Market pressure kept everyone honest.
The Technology Group was the R&D arm. They assessed Carrier’s tech profile continuously. A special group within them focused on service technology, like remote diagnosis. Every business unit funded this work.
Operational support teams worked as internal consultants. They were organized around throughput processes and had one integrated mission: reduce time cycles, eliminate waste, achieve flexibility, and hit total quality. All at once, not as separate projects.
Business Processes That Actually Work
Three business processes tied everything together.
The Decision System used interactive policy teams. Each team had at least three management levels: the manager, their boss, and their direct reports. These teams made policies, not operational decisions. Policies set the criteria that managers used to make their own calls. If no policy existed for a decision, the manager just decided. The team could create a policy later if needed. This is decentralization done right.
Performance Measurement was throughput-oriented, not functionally oriented. The measurement system recognized each unit’s contribution to other units, not just its own results. It included incentives for internal cooperation. They even considered differential tax rates for internal versus external transactions. The goal was to end the endless internal fights over pricing.
Target Costing was the financial backbone. Each unit’s budget was variable, tied to the total throughput of the system. Revenue for each unit was a percentage of total system throughput. Costs were actual costs. So if the whole system does well, everyone does well. If it doesn’t, everyone feels it. This aligns incentives across every level and every unit.
My Take on This Chapter
This chapter reads more like a consulting deliverable than a textbook chapter, and that’s what makes it valuable. You can see exactly how systems thinking principles translate into real organizational decisions.
The most interesting detail is buried in the introduction. This design was never implemented. Changes in the corporate environment and strategy shifts got in the way. Great designs still need the right conditions to succeed.
Still, a lot of companies today are struggling with the same problems Carrier faced in 1992. How do you centralize and decentralize at the same time? How do you keep 90 autonomous units moving in the same direction? How do you measure performance without creating internal warfare?
The answers are all here. The question is whether anyone has the courage to implement them.