The Build-Measure-Learn Feedback Loop
Creating Real Value by Testing Ideas
In a flash of inspiration, one of your team members has come up with a dazzling idea for a new customer service product. It's fun and innovative, and you know that it's got the potential to make your life easier. In your excitement, you feel certain that other people are going to love it just as much as you do.
Your enthusiasm is infectious. You sell the idea to your bosses and win over your colleagues. You spend months working behind closed doors to bring it to life. It takes a great deal of money and effort, but eventually you have a product that you're proud of.
Then you launch it – and it fails spectacularly. You're left reeling.
Developing products in isolation like this, without checking your prospective customers' needs, is a mistake that many businesses make. They press on enthusiastically but blindly, and their ideas often fail. Only 50 percent of U.S. startups, for example, survive beyond their fifth year.
The Build-Measure-Learn feedback loop is a technique that helps you to realize when you've got things wrong, before it's too late to turn initial failure into eventual success.
Understanding the Model
Build-Measure-Learn is one of the central principles of Lean Startup – a highly effective approach to startup development pioneered by Eric Ries. (You can listen to our podcast review of his 2011 book, ‘The Lean Startup,' here.)
Ries says that the job of a startup is to find a successful revenue model that can be developed with further investment. Build-Measure-Learn is a framework for establishing – and continuously improving – the effectiveness of new products, services and ideas quickly and cost-effectively.
In practice, the model involves a cycle of creating and testing hypotheses by building something small for potential customers to try, measuring their reactions, and learning from the results (see Figure 1, below). The aim is to continuously improve your offering so that you eventually deliver precisely what your customers want.
Figure 1 – The Build-Measure-Learn Feedback Loop
Build-Measure-Learn may sound simplistic, but it's been a game-changing technique for businesses that previously developed products without getting potential customers' input. Sometimes, companies would get lucky, but many wound up making sophisticated products that no one wanted.
Build-Measure-Learn improves on the "just do it" approach with an incremental, iterative methodology that replaces assumption with knowledge and certainty.
Build-Measure-Learn is not appropriate for a project that demands a low rate of failure and that has been successfully completed many times before. It best suits fast-changing, high-risk environments, where research is difficult to conduct and customers are unclear about their needs.
Using the Model
Follow these steps to work through a Build-Measure-Learn feedback loop.
Step 1: Plan
The model may be called "Build-Measure-Learn" but, if you follow that sequence and jump in at the "Build" phase, you'll be missing the mark. Instead, it's essential to start with a planning stage.
Your first task is to define the idea that you want to test and the information that you need to learn. You do this by developing a hypothesis – your prediction of what will happen during the experiment.
Your hypothesis could focus on anything from product features and customer service ideas to finding the best pricing strategies and distribution channels. You might, for example, hypothesize that "increasing the frequency of our newsletters from two to four per month will increase overall revenue."
Next, decide what you'll need to measure to test your hypothesis, and plan how you'll collect your data. Interviews, surveys, website analytics and specialized software programs are common methods for gathering data, and the BADIR process will help you to structure your study.
Step 2: Build
Your goal here is to create a Minimum Viable Product (MVP) – the smallest possible product that allows you to test your hypothesis.
It could be a working prototype or a basic advertisement or landing page. It could be a presentation slideshow, a mock brochure, a sample dataset, a storyboard, or a video that illustrates what you offer. Whatever MVP you choose, it needs to show just enough core features to attract the interest of early adopters – the people who'll likely want to buy your product as soon as it launches.
For example, the first 5,000 people who subscribed to the cloud-based file sharing company Dropbox™ did so before its service was launched. They'd been convinced by the strength of Dropbox's MVP – a 90-second video explaining the service that it was about to offer.
As you work through repeated iterations of Build-Measure-Learn, your MVP will become more complex. But your priority, as a startup, should be to prove the demand for your proposed product, not to build a fully functioning model that's full of advanced features.
Once you've done this, you can raise further finance to build a more richly featured product.
When you've created your MVP, launch it and collect data using the techniques that you chose in Step 1.
Step 3: Measure
Here, you measure the results that you obtained in Step 2. How does what actually happened compare with your hypothesis? Is there sufficient interest in your idea to continue developing it? Does the data show that you'll be able to build a sustainable business around your product or service?
Step 4: Learn
By the time you reach this stage, you'll be equipped to make sound, evidence-based business decisions about what to do next.
There are then two ways forward:
Persevere: Your hypothesis was correct, so you decide to press on with the same goals. You repeat the feedback loop to continuously improve and refine your idea.
(Even though your idea has achieved sufficient initial success to persevere with it, bear in mind that your next iteration may not do so. Be prepared to pivot in the future.)
Pivot: The experiment has refuted your hypothesis, but you've still gained valuable knowledge about what doesn't work. You can reset, or correct your course and repeat the loop, using what you've learned to test new hypotheses and carry out different experiments.
You can pivot in various ways. For example, you could develop a single feature from your MVP (called "zoom-in pivoting") or focus on a different type of customer ("customer segment pivoting"). Or, you could try delivering through a new channel ("channel pivoting") or use a single feature as the basis of a different product ("zoom-out pivoting").
Early enthusiasm for Dropbox's idea persuaded the organization to persevere. However, it made mistakes when attempting to expand its initial user base, so it had to pivot several times during subsequent iterations of the feedback loop.
Build-Measure-Learn often generates bad news, particularly during early cycles. You may need to pivot repeatedly before you can persevere. This is simply because you're testing at an early stage of development, before you fully understand what customers want and before your offering has any value-adding features.
Pivoting can be damaging to your morale, but remember that it's a critical part of the Build-Measure-Learn process. Every failed or underwhelming MVP is an opportunity to learn and grow, and to recommit to the feedback loop. Flexibility and having the courage to persevere are key to the success of Build-Measure-Learn – and of building a business generally.
This is where you need to think about your "runway" – the amount of money that you have available from investors. If your runway is short, you might only have time to try out a few ideas before you run out of funds. If your runway is longer, you have many more opportunities to pivot.
Eric Ries pioneered the idea of Build-Measure-Learn in his book, "The Lean Startup." It is a learning and feedback loop for establishing how effective a product, service or idea is, and doing this as quickly and cheaply as possible.
Follow these steps to use the Build-Measure-Learn feedback loop:
Step 1: Plan your experiment: learn, measure and build – including developing a formal hypothesis.
Step 2: Build a minimum viable product, and test it.
Step 3: Measure the results against your hypothesis to decide whether you can develop a viable business around your product.
Step 4: Learn from your results, and decide whether to persevere or pivot.
Then, cycle back to the beginning, and keep on going around the loop as you develop your product.
The biggest advantage of this technique is that it minimizes the risk and cost of creating products or services that no one wants, and helps you to "zero in" on something that customers will embrace.
This site teaches you the skills you need for a happy and successful career; and this is just one of many tools and resources that you'll find here at Mind Tools. Subscribe to our free newsletter, or join the Mind Tools Club and really supercharge your career!