Lean Startup is a discovery journey

Padrão dos Descobrimentos monument near Lisbon, Portugal, that celebrates the Portuguese Age of Discovery during the 15th and 16th centuries.

As someone who cares deeply about Lean Startup and actively tries to help drive its adoption in companies of all sizes through hands-on teaching, I often hear “We have done a Business Model Canvas” or “We have talked to our customers”. When digging deeper though, I quickly uncover that these were mostly one-off activities instead of a starting point of an ongoing journey. After a short introduction to Lean Startup usually through workshops or a “Lean day”, people quickly return the next day to what they were doing before: building out their solutions or selling them (prematurely). It is clear that the fundamental meaning of Lean Startup has not been understood (or taught well enough!):

Lean Startup is a discovery journey!

Or expressed more formally:

Lean Startup is the continuous and iterative process of challenging in the real world what you believe to be true about your project* and turning these assumptions into concrete facts.

Let’s clarify the key points (highlighted in bold above):

  1. “continuous and iterative process”: applying Lean Startup is an ongoing activity throughout a startup’s or product’s lifetime with the goal of improving at each step.
  2. “in the real-world”: you test your assumptions with real people who you deem to be prospects for your product and in their environment.
  3. “concrete facts”: you gather evidence, e.g. prospects’ email addresses, a letter of intent, money!!!, that proves that your assumptions are true (validation) or are not true (invalidation). The latter is usually the case when you cannot get any positive signals from your prospects.

So, how could you implement Lean Startup as continuous process in your startup or larger organisation and thus increase the chances of success for your project? For me, the most critical and probably hardest part is this:

Practicing Lean Startup requires self awareness, willingness to change and rigorous discipline.

In other words, you need to be able to recognise and remove your own bias towards your project (primarily towards the solution). You need to embrace new ways of thinking and step away from how things were. And lastly, you need to keep distractions to an absolute minimum and focus on one thing at a time.

Generally though, people with ideas (such as founders, corporate Innovation Managers etc.) tend to not want to get their ideas challenged or even proven incorrect. And why would they? They have relevant work experience, “know the industry” and may have invested considerable time and probably some money (either their own or someone else’s) into the project. Sometimes there is a functional prototype with some beta testers in place already or the VP of Sales and his team of “Inside Sales” reps are franticly trying to sell a solution to anyone who has come up in their LinkedIn search. Why stop now and consider putting your chosen customers, solution or revenue model into question? Because to build a sustainable business, you need to have a working business model: You need to able to repeatedly sign up a large enough group of customers, deliver value to them through a solution that solves their problems and then capture monetisable value from them, e.g. in form of sales or eyeballs.

So, once you have shifted your mindset, here is a suggestion on how to proceed — I am using an example for a new business idea but this process also works for new product features from step 3 onwards:

  1. Document your assumptions about your business idea (problem(s), customer segments, value proposition, revenue streams, costs etc.) using the one-page Lean Canvas**
  2. Identify the riskiest part(s) of your business model, i.e. those parts that absolutely need to be true or your business idea will fall apart.***
  3. Formulate a falsifiable hypothesis capturing what you want to test, who with, how long for and what outcome you expect from the experiment (usually a hard metric).****
  4. Design an experiment to test your hypothesis.*****
  5. Run the experiment — keep to the set timeline and don’t change anything yet in order not to influence the experiment
  6. Compare the experiment’s results with your expected outcome (metric) and, based on that outcome, make a decision to either:
  • Persevere (if hypothesis is validated) — continue with the next experiment to slowly validate/invalidate the other assumptions about your business model
  • Pivot (if hypothesis is invalidated) — investigate and understand why your hypothesis is not true, e.g. you set an unrealistic metric to start with or addressed the wrong customer segment etc.

7. Iterate based on the learning from the previous experiment

8. Define a new hypothesis

9. Design a new experiment

10. Run the next experiment

11. Review the outcome and make a decision on next steps

12. Rinse and repeat the above process until you have sufficient evidence (facts!) that your business model hypotheses are true or are not true.

Important caveats:

  1. This process is not as linear and smooth as it may suggest. It is a constant back and forth between developing hypotheses, testing them and finding evidence that validate or invalidate your assumptions.
  2. You don’t necessarily need to develop, i.e. code, anything to test your assumptions. There are plenty of affordable and easy-to-implement techniques and tools you can use for your experiments. See this link.
  3. You should be prepared to discontinue with your business idea if you are repeatedly getting negative signals from your experiments, e.g. your success metrics are not met. Quitting a business idea requires being honest towards yourself, your team and your project’s viability and the strength to admit that your business idea as you envision it today is not working.


* A “project” could be a new product feature for an existing product to a new business idea.

** Most people I talk to are aware of the Business Model Canvas and less so of the Lean Canvas which I personally prefer because it specifically calls out the customer problem(s) you are tackling. Read this article by Ash Maurya for a good explanation about the difference between the two canvasses.

*** For early-stage startups, the riskiest part usually evolves around the customer — problem pair, i.e. is the problem/need you identified real and big enough for the chosen customer

**** A falsifiable hypothesis can be proven incorrect because it is simple and unambiguous, measurable and it describes a relationship and is achievable. For practical examples of falsifiable hypotheses, check out this link.

***** A great tool for designing and running experiments is the “Experiment Board“. Be aware of the market and environment which you are in to choose the right experiment. See also this handy list of traps when experimenting and how to avoid them.

Originally published at fforward.me.