Abduction, Synthesis and Lean UX

What is Lean UX?

I was fortunate enough to be invited to a Lean UX workshop, hosted by 3M featuring Jeff Gothelf. Jeff is releasing a book soon and the first chapter is a free read. Check it out. In summary, Lean UX is built off many of the tenets of Lean Startup and agile development to build the right product for the right market.

One of the core practices that is supposed to make Lean UX a successful approach to product design is the generation of hypotheses and the subsequent testing of those hypotheses through experimentation/prototypes with real users in market. I've been following the Lean UX movement from afar for quite some time, largely because the experimental nature of it really resonates with my approach to work and life as a whole. I've always worked with a "move fast and break things" approach, which caused a lot of tension in my early career as a manufacturing engineer—it's usually not a good idea to move fast and break equipment that costs hundreds of thousands of dollars—but is very well suited for digital products. Lean UX provides a framework to move even faster and to break less things. That's the aim in any case.

After participating in the workshop I was driven to further investigate the concepts that Jeff was promoting. Fortunately for me, I went on a two-week vacation right after the workshop, giving me room to read and reflect. I have always been interested in the concepts of abduction and synthesis, but have always felt that I didn't fully comprehend them for myself so I got deep into some fairly heavy academic reading. I'm still very much in study mode, but was encouraged to share what I've learned so far.


Abduction is a somewhat difficult concept to grasp, not because of any cognitive complexity to it, but because it is so similar to induction. I find that I more easily understand complicated concepts by drawing clear contrasts and by creating a mental model using a spectrum to convey either end of the contrasting concepts. You can think of abduction and induction as being on either end of a certainty spectrum—from low certainty to high certainty—where they are both methods of inferential discovery. It all comes down to probabilities, where abduction is essentially drawing intelligent guesses or hypotheses about a given situation and induction is making a similar guess but with a higher probability or certainty (we can guess the sun will rise based on a lifetime of seeing the sun rise, a fairly certain theory of gravity, etc.).


If abduction is simply guessing in a way that makes you look like a boffin instead of a buffoon, how is it that this practice is supposed to lead to any successful discoveries or products? Enter synthesis. Abduction isn't about making wild guesses, it's about generating sound, contextually relevant hypotheses. Synthesis is the beginning of contextualizing your guess-work.

All knowing is inferring and that knowing is done through the comparison of different mental models. I can share a concrete example of this through some of my own experiences. As I stated above, I started my career as a manufacturing engineer in a lean manufacturing environment. Before I knew anything about UX or software development, I was doing Kaizen exercises with our machinists, setting up Kanban inventory systems, and doing blue yarn analyses for material flows. As my career trajectory started to move towards software consulting and product design I had to quickly adapt my training, experience, and education to a new context. But I didn't have to start from scratch, I had abduction and synthesis on my side and could make the connections between the mental model of manufacturing physical products and apply those lessons to digital products and workflows.

Conclusion // tl;dr

Jeffery Zeldman and Khoi Vinh made an important point about the importance of learning through writing. That's what I'm aiming to do with this post.

  • Lean UX is about creating a repeatable process for finding product/market fit.
  • Generate hypotheses and test those hypotheses with as short a feedback-loop as possible (aka Minimum Viable Product—MVP).
  • Use abductive reasoning and synthesis to make your hypotheses as contextually relevant as possible by comparing models you've experienced in the past that share similar attributes.
  • … profit.
Posted on January 5, 2013 and filed under career, design, Strategy, UX.