A Language and Process for Designing Businesses.
At the bottom of this piece are slides from two documents: A Systems Design Language [View/Download PDF] and an Evolutionary Design Process [View/Download PDF] that form the basis of a method for Venture Design.
Creativity: Pattern Recognition and Noticing
Creativity is about deliberately creating or spotting lucky accidents (going wide) and then editing them (going narrow). Going narrow can be partly logical and can use deductive reasoning, but going wide can’t. In reality it’s a logical fallacy to be able to deliberately do something by accident, so you have to put yourself in or create an environment that has a degree of experimentation, originality or noise to create these accidents and then notice them. Creative people don’t magic things out of thin air, they are good at noticing unusual things (sometimes in their own subconscious) and then recognizing patterns, so they can edit them together in a coherent way.
Design: Codified Creativity
Two of the differences between design and art is that a designer usually doesn’t make the end product and the process of making it often needs to be repeatable. The first difference requires codified instructions, for someone else to make something which might be new and original. The second requires a process to generate these instructions, repeatably, for multiple projects. Unlike art, all design requires a codified process for it to be scaleable. The best ones are seemingly trivial and can be represented as simple diagrams, so that they are easy to use. They work in that they condense something complex into a single picture.
Venture Design: Systems Consisting of Products +Business models.
Most designers design things (products), we design companies (products intertwined with business models), so I developed a simple process and language for Venture Design, with some obvious looking but hopefully useful diagrams that encapsulate them.
A Popular Design Process
The Double Diamond approach (go wide then narrow to articulate the problem to solve and then do it again to solve it) is a popular model for design process that’s nice and simple, but it doesn’t quite match either traditional or agile environments for creating things, in two important aspects:
- Traditional design like architecture focuses on the second diamond having already received a brief that defines the problem and adds a third to deliver instructions for how to build the end product.
2. Agile environments test things in the wild and evolve them over time.
Double Diamond Design Limitations.
In traditional processes such as architects use for designing buildings, there is an implicit third set of diamonds between the problem and the delivery of solution that consists of an idea and a plan to solve it (what is actually called the design in architecture). Architects don’t build buildings, they deliver designs to contractors and their delivery is more like a plan than a prototype. The prototype (a model) is usually a marketing or presentation tool. Apple’s hardware design is somewhat similar (Apple design hardware, they don’t make it), but where models are delivered to those who make the product. Whether there is a third, implicit set of triangles for Apple, depends on what you mean by ‘prototype’.
Modern, agile software design contrasts with traditional design, in that (a) ideas can be quickly mocked up to a minimum viable level and tested in the real world and (b) the research to define the brief can be considered much more part of the design process than in areas like architecture (where the brief is often handled by the client). The set of diamonds that represent the user research and those that are to do with delivering the product are both considered to be the design, and the middle set (the ideas and plans) are missing. This is the reason that Double Diamond (research and delivery) is such a useful shorthand, and quibbles about its accuracy are really about what is meant by ‘prototype’, however, the iterative nature of agile methodologies implies multiple iterations and therefore also implies multiple sets of diamonds. This kind of design relies on keeping going wide and narrow, to move things along — like a bowel movement, but with a better outcome!
There are more complex alternatives to the Double Diamond process that do try to account for design complexity by nesting, but they often lose the elegance of the original for what they gain in thoroughness.
An Evolutionary Design Process
For Anthemis Design which is about designing companies (Venture Design) I have developed a diagram that represents our process which basically says go wide and narrow multiple times, with feedback loops. The idea that this method of feedback loops is done in a way which is more like living systems is more than mere spin to reflect the way Anthemis works to nurture ecosystems, since designing companies is fundamentally different from designing products like tables and chairs. Companies aren’t static things, they are systems with inputs (capital and resources) and outputs (products and services) and therefore the process of designing them is different and much more like designing a process itself. A process for designing a process.
A Systems Design Language
The ‘evolutionary design process’ should be coupled with a second document — the ‘systems design language’. This treats human centric design as a particular type of system (where a company, a sector, a tribe or a whole ecosystem can be treated in the same way). The requirement of arrows for input and outputs are like paired transactions in accounting systems but used to represent any exchange of value, monetary or personal (or energy/entropy for that matter).
In accounting, double entries for transactions mark their beginning and end, so they can be represented as pairs rather than single values. Two values and the indication of one being a debit and the other a credit gives a line with a direction and so transactions are more like vectors than scalars — and therefore they can be represented by arrows. The paired arrows in our model go beyond double entry bookkeeping in that they record all value, whether monetary or not, and so record all transactions as double pairs (i.e. the credit and debit of value (monetary or otherwise) for both parties (e.g. buyer and seller)). This allows systems and ledgers to be connected together into one ‘balance sheet’ of value for the whole system, while accounting for the differences in perceived value between parties.
Because the type of businesses we design are Financial Services ones, the double entry system of value and money/energy (quantified value) exchange, as inputs or outputs are particularly important.
The systems approach has an important benefit. It allows you to marry human centric design (the approach design agencies tend to take) with awareness of markets as a whole (which is the approach entrepreneurs tend to use) so that anthemis can then have a design approach which takes best of the agencies like Ideo and those of entrepreneurs and investors within the startup world.
In order that the elements of the diagrams used in the language correspond to things currently used by both designers and entrepreneurs they map directly to both the Business Model Canvas and the Value Proposition Design Canvas. As sheets with boxes to fill in, these are more useful ways of collecting information. The language is designed to pull this data together to show that it fits a model which represents the mechanism of how a business works.
A Language and Process for Venture Design
By having a design ‘language’ to create a snapshot in time and a ‘process’ to show a design over time, you have the necessary documents to codify the way to design a system like a business.
One way of looking at it is that the systems design language is equivalent to a design version of a balance sheet (language/snapshot) and the evolutionary design process is equivalent to a P&L (process/over time) for how design is done specifically in the context of creating businesses.