In case you don’t know, I’m a big advocate of “Lean UX” and the work of Jeff Gothelf on this topic. It was Jeff who originally coined the term ‘Lean UX’ and wrote a great book about it, together with Josh Seiden. In short, Lean UX is all about not doing all your product design upfront, designing in more iterative fashion instead. I recently attended a talk by Jeff in London, where he talked about the concepts behind Lean UX, and here’s what I learnt:
- Cut out the design phase – Jeff stressed that the traditional design phase, where design is done in isolation, needs to go. I totally agree. Far too often designs are being thrown over an imaginary fence and implemented in a way that didn’t match the desired user experience, as intended by the designer. Instead, designers and developers should collaborate around ‘transient artefacts’, a simple sketch or prototype (see Fig. 1 below), in order to come to a shared understanding. These artefacts are by no the finished article, but create a shared understanding of what we’re looking to achieve and why.
- So what!? – The main idea about being ‘lean’ and ‘Lean UX’ is to “do less, more often.” Instead of just focusing on shipping features as often as you can, Jeff argued that the goal should be to achieve specific user outcomes as frequently as possible. I can imagine that this means a change in mindset and approach for some product teams. I believe, however, the ability to constantly ask the question “so what!?” from a user’s perspective is critical to building successful products. Whether you ship 2 or 50 features per iteration, this becomes totally irrelevant if you’re not changing customer behaviour.
- Humility and risk mitigation – I personally believe that as a product persona, you need to have a healthy degree of humility, and Jeff addressed ‘humility’ within the context of assumptions. He acknowledged how hard it can be for all of us to admit that we don’t know what the end state of a product or solution is going to look like, especially when you haven’t spent months creating a full product specification upfront. However, this is a risk that can be mitigated by being humble, acknowledging that you’re making assumptions and picking the riskiest assumptions to test first. In practice this will mean that you’ll be doing user research and getting feedback as part of each iteration.
- Change definition of “DONE” – Instead of just taking a feature release as the final stage of the definition of “DONE”, Jeff suggested that “changing customer behaviour” should be the ultimate definition of DONE. You pick a tactical metric that reflects the impact that the product is looking to make on customer behaviour. For example, you focus on a leading indicator to represent ‘customer intent’ or ‘conversion’ and something is only considered DONE if you’re moving the needle on your chosen metric (see Fig. 3 below).
Main learning point: I always learn something new from Jeff Gothelf and it felt no different with his recent talk in London. The main thing I took away from his talk was the importance of focusing on user outcomes instead and changes in user behaviour, as opposed to just shipping features.
Fig. 1 – Visual depiction of a transient artefact – Taken from: http://www.jeffgothelf.com/blog/you-dont-need-a-ux-portfolio/
Fig. 2 – Acknowledge that we’re making a lot of assumptions when building products – https://medium.com/the-job-to-be-done/replacing-the-user-story-with-the-job-story-af7cdee10c27
Fig. 3 – Build-Measure-Lean and Think-Make-Check – Taken from: http://www.smashingmagazine.com/2012/10/lean-startup-is-great-ux-packaging/
Related links for further learning: