Designing Business Architecture: The Multidimensional Framework
This is post 13 of 23 in a series on Systems Thinking: Managing Chaos and Complexity by Jamshid Gharajedaghi (ISBN: 978-0-7506-7973-2).
Previous: Defining the Problem - Telling the Story
Chapter 7 is where the book gets really hands-on. All the theory from the earlier chapters? This is where it turns into something you can actually build with. Gharajedaghi introduces a framework for designing business architecture, and the core idea is called multidimensional modular design.
Let me walk you through it.
Why One-Dimensional Thinking Fails
Most organizations pick one thing to organize around. They build everything around their product. Or their market. Or their technology. And for a while, that works.
But the competitive game keeps changing. When the market shifts, a product-focused company suddenly needs to reorganize around market segments. When technology leaps forward, a market-focused company scrambles to restructure around new capabilities.
Each of these switches is painful. People lose their roles. Teams get reshuffled. Knowledge gets scattered. And by the time you finish reorganizing, the game has changed again.
Gharajedaghi’s point is straightforward: if you keep locking yourself into one dimension, you’re setting yourself up for an endless cycle of restructuring.
The Three Dimensions of Business
Every business can be described along three dimensions:
- Technology (Inputs) - the know-how, the tools, the core competencies
- Products (Outputs) - the actual things you make and deliver
- Markets - the customers you serve and how you reach them
Traditionally, companies pick one of these as the boss and make the other two follow orders.
Apple is the classic cautionary tale. They had an incredible graphical interface operating system but used it only to sell Apple computers. The technology was way more valuable than the product it supported. Microsoft saw that gap and filled it with Windows.
When markets define the business, you get companies like Procter & Gamble: they make whatever they can profitably sell in supermarkets. When technology defines it, you get 3M, built around “sticking things together” and applying that expertise everywhere.
Each approach works until conditions change. And conditions always change. The answer is to treat all three as equals.
Multidimensional Modular Design
This is the big idea of the chapter. Instead of organizing around one dimension and subordinating the rest, you create an interactive architecture where product, market, and technology work as interdependent platforms.
Each platform hosts a set of modules. Think of modules as semi-autonomous units that can operate on their own but are designed to interact with modules on the other platforms.
The beauty of this design: you never have to reorganize again. When the competitive emphasis shifts from products to markets, you don’t tear up the org chart. You shift attention and resources between platforms. The structure stays intact.
Gharajedaghi frames this as the difference between “power over” and “power to do.” Traditional structures are obsessed with power over, with hierarchy and control. Modular design is about power to do, about giving each unit enough autonomy to actually get things done while keeping them connected to the whole.
That distinction hit me pretty hard. How many organizations do you know where most of the energy goes into controlling people rather than enabling them?
But First: Boundaries and Environment
Before you can design anything, you need to know what you’re designing and what you’re not. That means defining the system’s boundary.
The boundary isn’t a physical line. It’s defined by your stakeholders and what you can actually influence. Everything you can control or significantly influence is inside the boundary. Everything that affects you but you can’t control is the environment.
Here’s the interesting part about stakeholders. Stake and influence rarely line up. Customers have enormous influence (they can just stop buying) but low personal stake in your survival. Employees have massive personal stake but often low influence. Shareholders have huge influence but the least at stake, because they can just move their money somewhere else.
Purpose: It Depends on Your Paradigm
What is an organization for? The answer changes depending on how you think about organizations.
If you see an organization as a machine (the mechanistic view), it’s a tool owned by someone. Its purpose is to serve the owner. You measure it by efficiency and reliability.
If you see it as a living organism (the biological view), its purpose is survival. You measure it by growth. Profit becomes the fuel for growing, which gives it social value beyond just enriching the owner.
If you see it as a voluntary association of purposeful people (the sociocultural view), its purpose is to serve both its members and its environment. You measure it by doing more with less.
Gharajedaghi quotes Milton Friedman: “the business of a business is business.” But how you interpret that statement depends entirely on which paradigm you’re operating from.
This section also introduces the business model: how you generate value, package it, and exchange it for money. He uses Google as an example. Free service for users, money from advertisers. A business model that would have sounded ridiculous to anyone in the 1990s.
Functions: Whose Problem Are You Solving?
Once you know your purpose, you need to define your functions. The first question is deceptively simple: whose problem are you trying to solve?
Gharajedaghi has this great line. A true customer is “a pain in the neck.” They have something you want (money), and their satisfaction is the price you pay to get it. That’s refreshingly honest.
Selecting a product-market niche means answering four questions: Whose problem? What solution? How do we reach them? Can they pay?
The traditional approach assumes a bell curve of customers, where 90% are in the middle and you design everything for that majority. The “nerds” on the edges get ignored.
But Gharajedaghi points out the curve is flattening. The big middle is shrinking. The niche groups on the edges are growing fast. For new players entering a market, those niches are often the best opportunity, because the established players are too busy fighting over the shrinking center.
If you’re building a startup or launching anything new, this insight is gold. Don’t fight the giants for the mainstream. Find the underserved edges.
Structure: Outputs, Inputs, and Markets
With purpose and functions defined, you can now build the structure. Here’s how the three platforms work.
Outputs Dimension. This is where the organization’s mission gets accomplished. Output modules are semi-autonomous units responsible for a product, project, or program. Each one controls enough resources to be accountable for its own success or failure. Ideally, a product module is entrepreneurial: development, design, marketing, and profitability. But it should not own capital-heavy production facilities. Tying a division’s fate to a single product’s lifecycle is a death trap.
Inputs Dimension. These are the shared services and core technologies. Manufacturing, specialized functions, shared infrastructure. Input modules exist because some things are too expensive or too important to duplicate across every output unit. They earn their keep by charging market price for their services. This keeps them sharp and competitive rather than turning into bloated internal monopolies.
One warning Gharajedaghi gives here: never let a service function become a control function. When shared services start acting like bosses, output units will duplicate those services out of self-defense. You end up with worse service and weaker control. Keep service and control separate.
Markets Dimension. This is the interface with customers. It handles two things: distribution (representing the organization outside) and advocacy (sensing what customers actually want and bringing that perspective inside). The advocacy role is key. Someone needs to be the voice of the customer inside the company, and it shouldn’t be the same team trying to hit quarterly sales numbers.
Together, these three platforms form an interactive whole. They’re always in conversation, always adjusting, always redesigning how they relate to each other. That’s what makes the system adaptive.
My Take
The multidimensional modular design feels surprisingly modern for a book from 2006. It’s essentially what companies like Spotify and Amazon have been trying to build with their team structures, autonomous units that own outcomes while sharing platforms and capabilities.
The insight about power-to-do versus power-over is the kind of thing that should be on a poster in every executive’s office. Most organizational dysfunction I’ve seen comes from structures designed for control rather than capability.
The next post will cover the second half of Chapter 7: processes, internal market economy, and how the whole thing actually runs day to day.
Next: Designing Business Architecture - Processes and Realization