THE BUILD-MEASURE-LEARN CYCLE
Excerpt from How The Build-Measure-Learn Cycle Really Works | Cleverism
The Build-Measure-Learn cycle is a feedback loop that is said to be one of the core components of the Lean Startup methodology. Its goal is to turn uncertainties, assumptions and risks into knowledge or “sure things” that will eventually guide organizations and business towards progress. Through this process, the key unknowns can actually be transformed into knowledge that the startup can use in its product development – and business operations, as a whole. This whole process can also be called an experiment.
In order to understand how the Build-Measure-Learn cycle works, let us first take a look at its components.
Phases of the Build-Measure-Learn Cycle
There are three phases to this loop: Build, Measure, and Learn. However, it does not have to be strictly in that order. It is, after all, a cycle. In the book entitled “The Lean Startup”, Eric Ries said that, although the startup’s activities happen in the build, measure, and learn order, planning actually works the other way around.
In short, the startup will have to realize what it needs to learn, and then figure out what needs to be measured through the use of innovation accounting. This will eventually reveal whether the startup has gained validated learning or not. Finally, it will then figure out what product must be built.
Let us go through the phases of this loop. Before you can start, however, there is a need to “frame” the experiment. This is where you will identify what problem needs to be solved, and how you will run the whole experiment. This entails gathering data and detailing the background of the experiment, formulating a hypothesis and considering the variables of the environment. From there, you can now start with the feedback loop.
Phase #1. BUILD
In this phase, the startup’s goal is to build or develop its MVP – “minimum viable product”, or the bare minimum product that can be built for the purpose of testing a number of assumptions, or the hypothesis formulated – as quickly as possible. Before it can do that, however, the startup must first figure out what the problem that needs solving is.
Design the experiment. First, you have to build out the details of the experiment and figure out how everything will fit and mesh together. For this, you must have a solid and reliable method of collecting data, meaning the data gathered must be reliable and actionable.
Build the experiment. In this stage, think simple and small. Many startups tend to go big and complicated on the get-go, and they end up being overwhelmed and unable to handle it. It would be better to build the smallest possible increment that will still be enough for you to use to validate or reject the hypothesis you have made previously.
Run the experiment. This is where you will collect the data. The most common methods include conducting interviews or distributing questionnaires. In some instances, others may come out with prototypes for testing.
Among the activities that Eric Ries identified as part of the Build phase includes conducting unit tests, usability tests, refactoring, and cloud computing.
Phase #2: MEASURE
In this second phase, the startup must then determine whether real progress is being made or not, and this involves measuring the results obtained from the experiment performed during the BUILD phase.
Data analysis. Analyze the data obtained from the experiment. What happened? What are the implications of the data to your hypothesis? Make a comparison on what you hypothesized to what actually happene
Data organization. Organize your data in a way that will make it easily understood, and for the whole scenario to be easily comprehended by whoever listens to it.
Data Presentation. Make your presentation of the data as compelling as possible. You want the members of the organization or the company to be engaged, so make sure you present it in a way that will truly grab their attention and hook them.
In order to speed up measuring, Eric Ries suggested conducting activities such as split tests, real-time monitoring, funnel analysis, cohort analysis and search engine marketing, to name a few.
Phase #3: LEARN
This is where the startup will have to make a decision based on the measurements accumulated: should it “persevere”, or should it “pivot”? Persevere, in this context, means carrying on with the same goals, while pivot entails changing or shifting some, or all, of the aspects of the product strategy. Afterward, you would have to document your findings and share them.
The questions that are to be asked in this phase include looking into the knowledge that has been obtained. How should that knowledge be preserved?
More importantly, what are the next steps that should be taken by the startup?
Ries cited several activities for this phase including, but not limited to, conducting customer interviews, split tests, customer deployment, and smoke tests.
The 3 phases of the cycle can be simplified in the following activities.
Ask whether the new idea of the startup actually solves a problem for the users.
Quickly come up with an action or a program that will test the idea with the users. Perform reassessment or reevaluation if needed.
Obtain feedback from your us Focus on getting information that is relevant and will be useful in helping you create the product / service that is wanted or needed by the users.
Consider the sustainability of the product or service. Will you be able to maintain the current level of engagement or service?
Illustrative Example of the Application of the Build-Measure-Learn Cycle
For purposes of discussion, let us take a look at a mobile app startup and how it should employ the Build-Measure-Learn feedback loop.
Some mobile app startups would follow this process: they have an idea for an app, and they immediately write everything down on their business plan. Next, they design the app on paper, and they would then start writing the code and finish the development of the app. Once it has been finished, they will begin talking to customers and selling the app.
That is not how the Build-Measure-Learn approach works.
Application of this cycle would follow entail the following steps.
The mobile app startup comes up with an idea.
The startup comes up with a business model canvas, presented in a short business plan. (BUILD)
They start talking with customers and ask them what features they are looking for in an app. (MEASURE)
They will obtain the feedback of customers. (LEARN)
Using the feedback obtained, they will have to repeat Step 2 and make revisions on the business plan until they get it right.
Once they got the business plan right, they would proceed to the implementation of a prototype, or the MVP (minimum viable product) for testing. (BUILD)
The prototype is then shown to the customers. (MEASURE)
Feedback from customers is obtained and learned. (LEARN)
Step 6 is repeated, making improvements on the prototype until they got the app right.
THE ZAPPOS STORY
Zappos.com has been cited by many analysts as a great example of a company that made use of the Lean Startup Method, particularly the Build-Measure-Learn cycle.
Now considered to be one of the biggest online shoe and clothing shops in the world, Zappos was founded in 1999 as a startup. It was originally just a site where customers can order shoes. It is not a manufacturer or a retailer since it does not have any shoe inventory. This was the Build phase.
When the first orders started coming in, the founders of the startup went to a local store to purchase the shoes that have been ordered and subsequently shipped them to the customers who placed the order. This was where they implemented the Measure phase of the cycle.
The Learn phase came in as the founders tried to gauge the interest of the customers in their idea and even obtained their thoughts and reactions to it. Using the lessons learned from the first orders, they made the necessary adjustments and iterations, applied improvements, and continued the cycle.
Some could say that this has an element of “trial and error”, and they wouldn’t be wrong.
A crucial element in the whole Build-Measure-Learn cycle is the reliance on customer feedback and reactions. It can be said that much of the success of the application of this cycle in the case of Zappos was the fact that the online shoe store makes use of a loyalty business model and utilized relationship marketing heavily.
Zappos could have gone another route entirely. It could have started its shoe inventory and simple resold them. However, by using the Build-Measure-Learn approach, it was able to test the waters first, so to speak. Instead of spending a lot of time building its inventory and hiring personnel to handle operations, it stuck to just the founders doing all the work, and getting orders first before procuring the shoes. Their first attempt worked, and it was only then that the founders decided to continue (“persevere”) with it. Today, Zappos is no longer limited to the sale of shoes, but is also selling other items as well, and that could be attributed to them “learning” that customers are also interested in buying merchandise other than shoes or footwear.
That is not to say that Zappos would have failed if it did not use the Build-Measure-Learn feedback loop. It would probably still have succeeded, but they may have taken more time to achieve the results they wanted. What this feedback loop did for Zappos, however, was to make sure that the process was accomplished quicker. Otherwise, Zappos might have taken longer to become established as the successful marketplace and retail giant that it is today.
To summarize, the Build-Measure-Learn loop pertains to the cyclical process of turning ideas into products, measuring the reactions, response, and behaviors of the customers against the products that have been built, and learning whether to persevere or pivot the idea.