The Complexity Budget
Every organization has a limit to how much complexity it can absorb before it starts to fail. How an organisation spends its complexity budget defines it.
Every product or service is an offer to take responsibility for some sort of complexity on behalf of a customer.
Every business is an attempt to collectively manage some sort of complexity more effectively or efficiently than its customers can.
This is easiest to see with products or services whose value proposition focuses on how complex the task is. The industry that got me thinking about this was financial services. Financial services firms are often in the business of risk, and they earn their profit by managing risk more cheaply or more effectively than their customers can.
But what does it mean to manage risk? It means hedging against a dynamic and hard-to-predict system or environment - a system or environment that is too complex to manage directly. An insurer allows its customers to engage with that system or environment, and the benefits that flow from that, by taking responsibility for some of the complexity.
This idea is not just relevant to risk-based businesses, though. Literally every product or service handles some sort of complexity, no matter how simple that product seems to be.
A toaster handles the complexity of a safe, efficient, electrically-powered heating element - a task that is famously difficult for an individual to manage.
Toothpaste handles the chemistry and biology of tooth health.
A pen handles the complexity of a reliable and smooth flow of ink.
A car handles the complexity of safe motion at higher-than-human speed on a road.
Even the road itself handles most of the complexity of navigation. I don’t need to know about how the terrain changes between here and my destination. I don’t need to assess all the various paths I could take between those points. That complexity has been “screened off” for me by the road. The fact that I just need to follow the signs is a testament to the road’s capacity to handle the rest.
Types of complexity that products and services can address include:
Material complexity - the complexity of materials or component construction (as in the toaster), including geographical features (as in the road)
Social complexity - the complexity of coordinating and aligning interests and resources (deal-making, dating apps and recommendation systems)
Temporal complexity - the complexity of change, uncertainty and risk over time
Compliance complexity - the complexity of interpreting and complying with laws or regulations
Cognitive complexity - the complexity of providing usable human interfaces for non-human systems, or tasks beyond human scale (a GUI, but also an admin assistant)
Ecosystem complexity - the complexity of interdependent systems
Note, not all of this complexity is handled by technological innovation - traditionally, a great deal is handled by bureaucratic systems, which are made of people + technology (eg bureaucrats and writing/data management technologies).
Addressing Complexity: Compute & Context
Overall, products and services address complexity through either compute or context - where compute is the selection and transformation of inputs into outputs, and context refers to the conditions under which this transformation can be reliably executed.
In addition, value seems to be proportional to the scale of complexity managed. A commissioned painting is worth more than an IKEA print, not because the output is more complex, but because the inputs (the client’s brief) are more complex, and so transforming those into an acceptable output is more complex.
By contrast, a pen that works in space executes the same transformation of inputs (hand motions) into outputs (glyphs on page) as the terrestrial version, but the space pen is worth 10,000x the pen that only works on earth, because it manages the complexity of a huge variation in context (in this case, physical conditions).
However, as the Russians discovered during the space race, while space-pens are incredibly hard to create, all pencils work in space, which tells us that not all products have the same sensitivity to context, AND if they address the same job-to-be-done (JTBD), the one with the least sensitivity to context will often win (delivering similar quality, but with decreased risk, or reliability across a broader context).
No wonder designers obsess about understanding the JTBD first. It’s very easy to add unnecessary complexity to the compute or context required. This is also why innovation can come from simplifying the job to be done - in this case, simplifying the JTBD from “create ink glyphs in space” to “create glyphs in space”.
An app with a better interface is worth more, because it reduces the relative complexity of inputs to outputs - offloading the complexity from the user to the designer and through them to system.
Complexity Budgets & Differentiation
Every business has a limit to the complexity it can and will handle on behalf of its customers - and this impacts both the promises an organisation can make (in the form of products and services) and their success at fulfilling those promises. Let's call this a complexity budget.
No business actually succeeds in taking on total responsibility for the complexity that a customer needs addressing. Contracts and terms & conditions exist to make these limits explicit, but I want to focus instead on the choices that we make about the kind of complexity each business chooses to service, and the kind they decline or externalise in some way.
Businesses differentiate themselves by spending that overall complexity budget in different ways across their value chain
Consider the value chain of two different restaurants - a fast-food joint, and an expensive high-end restaurant.
Both of these can be profitable (or unprofitable) businesses. What differentiates them is not success, but what sort of complexity they embrace through the job they offer to do for their customers, and how that complexity is distributed across the value chain.
A high end restaurant will bias toward spending its complexity budget on the dishes themselves, on the sourcing of high quality produce, and the quality of the service experience.
A fast-food restaurant will bias toward spending its complexity budget on the menu (esp variety of dishes), while constraining the complexity of the dishes, and the input cost to the customer.
For a fast food restaurant, especially a chain, much of the complexity lies in delivering consistency of product, despite variations in staff, suppliers and location-specific logistics. This is a kind of complexity of context.
The easiest lever to pull to reduce complexity in both cases is the menu. A degustation menu is a way of reducing the complexity of operations in the kitchen (fewer combinations to service), and a way of reducing the complexity of the ordering process.
Supplier businesses spring up to lower the complexity burden for the restaurant itself. Provedores handle the complexity of sourcing high quality produce, and often the complexity of prep work, by providing pre-prepared ingredients.
Innovation & Transformation
When viewed from the perspective of a complexity budget, we can start to distinguish some terms that often get confused.
Innovation - some invention that changes the complexity budget of an organisation
Transformation - a change in the distribution of complexity across the organisation’s value chain
When Steve Jobs radically simplified Apple’s product suite on his return as the company’s CEO in 1997 after the acquisition of NeXT, he had diagnosed the company as being incapable of managing the complexity it was offering to handle for its customers.
In response, Jobs not only reduced the complexity of the value chain, he also redistributed it, prioritising product performance (via design and ease of use) and brand. In other words, he sacrificed addressing the complexity of the market “context” (a broader set of customer needs across many personas), in favour of delivering better “compute” to a simplified market. This was primarily a transformation.
Similarly, when businesses expand into new geographies or markets, they promise to accommodate a new level of complexity of context, either through adapting their products to those new markets, or providing consistency of product in a different environment.
This increase in complexity means that the shift from national to multinational organisations implies creating or leveraging an innovation of some kind, and multinational business practices are a kind of socio-technical innovation.
Innovations do not determine transformations
If we think of innovations as increasing a complexity budget, especially a multi-affordance technology like AI, it’s easy to see why there isn’t an answer to the question - how will this change business?
This is because the additional complexity budget that such innovations provide can be spent in multiple ways - it can be spent on increasing the quality of the product, increasing the variety of products, increasing the degree of personalisation in the selling, manufacturing or servicing of the product. What matters is the kind of distribution of complexity that the business will transform towards.
The internet reduced the marginal cost of delivery of digital content to almost zero. This allowed new business models to be developed that focused on the complexity of curation, over the complexity of distribution, like Netflix or Spotify. But even among these, there is a variation in the complexity of content preparation (eg Apple music’s lossless compression vs Spotify’s lossy compression), which is akin to the difference between a chicken sous-vide and a chicken sandwich.
The point being that the distribution of complexity across a value chain is a design choice.
Value and Progress
If we now agree that every product or service is an offer to take responsibility for some sort of complexity on behalf of a customer then a story emerges about both value and progress.
Value lies in the transfer of complexity from one entity to another, from an individual to a business, or from one business to another.
Meanwhile, progress is an increase in the level of complexity we can afford to manage safely and effectively as a community.
So if every organisation and every community has a complexity budget — and value and progress are a function of how we spend it — then perhaps the real challenge isn't how to reduce complexity, but how to spend it well.
This raises questions at multiple levels.
A business executive might ask:
Are we taking responsibility for the right mix of complexity?
Which sorts of complexity are we avoiding, that might actually be worth taking on?
Which parts of our value chain are under-utilised - could handle greater complexity - and which are overburdened?
A designer or team leader might ask:
Where is the complexity debt in the organisation, and how is it being paid? What complexity is explicitly handled by process or system, and what is handled by our people, invisibly, through their situational adaptability and creativity?
Complexity budgets and ethos design
These aren’t just strategic or operational questions. There are ethical stakes involved - questions that go to the heart of what I call Ethos Design - designing practices that enable people to act responsibly in complex, real-world situations. This is because the idea of a complexity budget (as a natural limit) provides a bridge between value chains or business model design and responsibility as the capacity to handle complexity with care.
A company that offloads promised complexity onto its users or the public is ethically misaligned. So is a company that expects its employees to handle complexity beyond their capacity.
A well-designed product isn’t just elegant - it’s attentive to and considerate of who’s carrying the complexity load. Those design decisions are themselves ethical decisions.
What do you think?
This piece is part of a broader inquiry into Ethos Design - how we shape environments, systems, and tools to support ethical agency in complex work. If this resonates, challenges, or intersects with your experience, I’d love to hear from you. What are you seeing in your own context? Where do these ideas meet (or miss) the realities you work in? For example:
What other kinds of complexity have you seen products or services absorb on your behalf? What sort of complexity gets neglected?
What sort of complexity do you grapple with in your own role or practice? Can you interpret your own personal role-crafting in terms of complexity you handle, delegate or exclude?
Where have you seen a team or business quietly exceed their complexity budget? What does it feel like when you are implicitly expected to handle the gap?
Have you experienced a transformation that didn’t begin with an innovation - but with the redistribution of complexity across the business, or across it’s external relationships with customers, suppliers or partners? Have you seen this done successfully, for the benefit of all?