Every project that I have worked on, especially in corporations innovating, have each had their own interpretation about what iterating and learning means. Discussions around when to apply the cycle of build-measure-learn, how to apply it and most of all, when and what to start building, often conclude in the building of the product itself. This is potentially the most wasteful place to start, and yet very hard to resist because of the deceptively encouraging sense of progress provided to management and the team. To be fair, I completely empathise with the questions and confusions around the cycle. How is anyone to know, that in the context of learning under great uncertainty, build refers to anything that is done at that stage to best learn about the greatest uncertainties and risks. The more cost and resource efficient it is, the better it is for learning. Keeping it light and lean creates conditions conducive to a sharp learning curve and a willingness to be wrong and re-run the cycle.
Using a real-life example, let’s go through the phases of uncertainty to understand what to build in order to learn enough for the next stage. We start right at the very beginning – when all there is, is a broad vague topic. Something to the effect of ‘we want to do something new and profitable in a particular space’, e.g. Handyman tools. We don’t yet know, what kind of tools, to work on what kind of materials, to build what kind of structures, to serve what kind of customers. We don’t even know what problem we are setting out to solve. And that is all ok, just as long as there is a methodical acknowledgement of the known unknowns. Just as long as we don’t go out and build that proverbial hammer!
Hypothesis: [None]
Topic identification stage:
Exploring the behaviour and challenges of working with handyman tools.
Build: Craft effective experiments or methods to conduct generative market research (e.g. Customer discovery).
Measure: Different kinds of problems. How many problems there are, and how many of them are worth solving.
Learn: Frame problems related to working with handyman tools and who it is that’s facing them.
Hypothesis: [Vague] A lot of professional handymen experience hindered mobility with drilling machines. If the drilling machine would be less cumbersome and unwieldy, they could access difficult areas on construction sites better.
Problem validation stage: How serious this problem is for Handyman Joseph in the south of Munich. How high his motivation is to solve it.
Build: Craft effective research methods and experimentation to learn why the problem effects Joseph and what a solution would be for him (generative product research and evaluative market experiments e.g. Contextual inquiry / ethnography, landing pages, competitor usability).
Measure: Impact of reduced mobility for Joseph on construction sites. How often he faces this problem, how many potential Joseph’s exist.
Learn: What is the benefit in solving the problem, what the potential market size, what potential solutions are already available.
Hypothesis: [Specific] if Handyman Joseph would have a solution to the cabling, he would be able to access difficult areas on the construction site.
Between a lot of swearing, handyman Joseph is constantly challenged with the cable on his drilling machine. Invariably getting caught in obstacles, it is a bother to roll up and is never long enough, making it hard or impossible to access some areas of the construction site.
Ideation phase: Go through possible solutions to solve Joseph’s cable problem.
Build: Build quickest low-key testable solutions for Joseph to try out, while on construction site (using evaluative product research e.g. A/B testing, paper prototypes, mock-ups, low-fi prototypes, wizard of oz).
Measure: The degree of benefit of each approach for Joseph.
Learn: Which ideas showed promise. What further ideas/solutions were derived or evolved from the ones that were tested.
Hypothesis: [more specific] If Joseph’s drilling machine didn’t have any cabling, it would not
only increase his mobility and ability to work on hard to access areas, but also reduce his risk of accidents through cable entanglement and get rid of issues caused by cable wear and tear, like shaky contacts.
Solution iteration phase: Start product development to evolve and refine solution to best fit problem.
Build: Build iterations of testable solution for Joseph to try out while on construction site (using evaluative product e.g. A/B Testing, usability tests, iteratively functioning products).
Measure: The degree of benefit for Joseph.
Learn: Improvements on solution to arrive at the best marketing product for Joseph type customers.
Hypothesis: [Crystal clear] If Joseph’s drilling machine is equipped with a battery pack that allows unobstructed drilling for an entire day of work, he can work independent of electrical power on construction site, has no cable related issues like shaky contacts or safety hazards due to entanglement.
Product development and launch phase: Continue product development with constant feedback loop from customer.
Build: Iterate through product development and features to refine product.
Measure: The degree of adoption and satisfaction for Joseph. (using evaluative product experiments e.g. Usability tests, functioning products, analytics/dashboards).
Learn: Improvements on solution to arrive at the best marketing product for Joseph type customers. Expansion of existing solution to other areas.
Rather than building and selling the proverbial hammer, carefully crafted experiments on the basis of their potential learning value, led to the evolution of the portable drilling machine with replaceable sets of chargeable batteries, thus creating a solution applicable and scalable to a range of tools, in industry and domestic grade.
Experimentation, as it goes, requires rigorous discipline. Organisations that embrace experimentation don't pretend to know all the answers up front and are comfortable with uncertainty and ambiguity, experimenting to learn rather than to produce an immediately marketable product or service.
There are very effective methods applicable in each stage of uncertainty, depending on where the initiative is and what would instruct the next step best. Choosing the right activity, the right experiment for that stage, moving with speed not only affects where the team is headed, but could differentiate successes from failures.
For a good understanding of the different kinds research and experimentation with a nicely compiled list by @TriKro, go here. Happy experimenting, happy researching!
Read also: