There is no customer — the customer is a complex network (and how it relates to experimentation)
Every customer engagement is unpredictable as it is by its nature a chain of unknowable actions and reactions that might come from you, anyone or anything else.
The first thing that happens in an engagement therefore is change, and any plan you had before it started is solving for a situation that no longer exists.
Given this unpredictable interconnectedness of actors, actions and reactions customer problems aren’t simple or complicated they are complex. And complex problems need different mental models, tools and designs to be solved successfully.
This is not a controversial perspective, as Andrew Dolkart comments on Modernist architect Mies Van der Rohe (Kevin Slavin, 1):
Mies understood that the geometry of his building would be perfect until people got involved. Once people moved in, they would be putting ornamental things along the window sills, they would be hanging all different kinds of curtains, and it would destroy the geometry.(1)
But, Mies opted for control, to treat the problem as simple — something that could be designed away.
So there are no window sills; there is no place for you to put plants on the window. He supplied every single office with curtains, and all the curtains are exactly the same.(1)
The reaction to this, according to Kevin Slavin(1) came with the Internet and User Experience Design (early 1990). Where more attention was put on all the facets of use and interaction. But the notion was still that it would be predictable, controllable and that the customers’ engagements would be imaginable. User Experience Design therefore sees the world as complicated.
30 years later the Internet is no longer individual digital buildings or online architecture, but everyday life. The Internet is indistinguishable from how or what we do to accomplish tasks, work, live, life, love.
Kevin Slavin(1) argues that today we need to relinquish control, we need to design for participation:
“This returns to the drivetime question about the designers of complex adaptive systems: Price was designing not for the uses he wished to see, but for all the uses he couldn’t imagine. This demands the ability to engage with the people in the building as participants, to see their desires and fears, and then to build contexts to address them.“
Again:
“To design for all the uses he couldn’t imagine”
And with this the customer is recognized as a complex problem.
Dave Snowdens Cynefin framework is a helpful mental model to understand how the customer has gone from simple, to complicated to complex.
In the model he points out that there are four decision making domains: simple, complicated, complex and chaos. And that each of these have different methods of problem solving fit to what these problems look like.
In the video above Snowden introduces these domains:
Simple problem: categorization and best practice (see the problem and put it into a box you are familiar with)
Complicated problem: analyze and good practice (expertise is needed to understand, there is not one but several ways of solving the problem)
Complex problems: experiment, amplify or dampen (apply safe-to-fail experiments instead of fail-safe designs)
Chaos: According to Snowden, if you find yourself in the chaotic domain you are either there deliberately (for the purpose of innovation) or by mistake. The difference between chaos and complex is that in chaos you are trying to get out, so you act instead of probe. Meaning the intention isn’t to learn but to stabilize.
The less we know about the customer the simpler we think our problems are (we are either Mies or see our customers as Users).
And the more we know about the customer the more unknowable and unpredictable we understand they are. And that we can’t control our customers we can only understand them and hope to influence them.
And this is why we use experimentation. Not because the customer problem is simple or complicated, but because (or when) it is complex. And we don’t use experimentation primarily to validate our theories and establish them as facts, but to learn. And to keep learning and keep learning still.
A complex system doesn’t stop moving and we need to keep up. The way to do this is not control, but response.
Resources:
(1) Design as Participation, MIT Journal of Design and Science
, Kevin Slavin, https://jods.mitpress.mit.edu/pub/design-as-participation
(2) The Cynefin Framework, Cognitive Edge, Dave Snowden, https://youtu.be/N7oz366X0-8
