Pin Down your Startup Hypotheses

As all you really have at the beginning of your startup are a bunch of educated guesses, your highest priority is to figure out what you’re guessing wrong. In other words: can you build a viable business around your idea?

Those coming from an industry will start from verified market insights. However, by definition, startup is a learning organization, searching for a repeatable business model. Unless you are repeating something previously proven, you can never be sure that you have all the pieces right. Coming from a consultancy business myself, and knowing the pain of my customers well, I was really surprised by the assumptions I was making once I put them on paper and examined them with my customers. 

The first step in validating your guesses is to make them explicit, or put fancier: document your hypotheses. I’ll give you some ideas on how you can do this.

In his book Four Steps to Epiphany, Steve Blank recommends you write a number of briefs, each documenting a different hypothesis (about the product, customers, channels, market, etc). Each is described in a subsection of the chapter “Customer Discovery” (the first of the four steps of customer development). They provide useful questions to ask yourself.

But there’s an easier way.

The easier way: business model canvas. If you haven’t heard or used Osterwalder’s business model canvas, I highly recommend to learn more about it (the first 72 pages of Osterwalder’s best selling book are free to download from his website).

What’s cool about the business model canvas is that it collects 9 essential building blocks of a business on a single sheet of paper. Miss one and you got no business. Go on, fill one out. These are your hypotheses.

In addition to being a lean way to document your hypotheses, business model canvas is also a great brainstorm tool. Print the canvas on a big sheet, and brainstorm around it with your team.

Business Model Canvas, 1 page summary of your business

Business Model Canvas, an easy way to document your hypotheses

What to fill in? Steve Blank ran a hands-on customer development course with his Stanford students: the Lean Launchpad. The students had to form teams around concrete ideas, and actually perform customer development of their idea during the curriculum. To do so, they had to get out of the building and talk to actual customers. The students used the business model canvas as a score card, and blogged about their lessons learned in interviews.

What’s cool about this course is the bunch of resources and examples it provides, all free and well documented on Steve’s blog. Each post in the Lean Launchpad series zooms in on a different aspect of business model providing.

  • A theoretical checklist for the specific business model aspect
  • Lessons learned: you get to see how the students put this theory in practice

I collected them for you below.

Business Model Checklists Lessons learned
Value proposition checklist, a useful deck urging us to think about the problem statement, technology/market insight (why is this problem so hard to solve), competition and the (minimum viable) product. Lessons students learned by testing value proposition hypotheses
Customer checklist: different types of customers and markets, multisided platforms, different types of customer roles and customer problems. Lessons students learned by testing customer hypotheses
Customer relationship checklist: it’s about demand creation. Lessons students learned by testing customer relationship hypotheses
Channel checklist: direct and indirect sales channels, channels as customers, differences between physical and digital products and channels. Lessons students learned by testing channel hypotheses
Revenue model checklist: what are the customers paying for and how much can they pay, how will you package and price your product, different types of revenue streams. Lessons students learned by testing revenue model hypotheses
Bonus material: a slide deck on 7 different SaaS revenue streams (skip to slide 27 for the juice),  a free e-book on software pricing Don’t Just roll the Dice (disclaimer: still on my to-read list).
Key partners, key resources and expenses: two separate slide decks, the latter by Osterwalder. Lessons students learned by testing key resources and expenses hypotheses

Helpful alternative geared towards customer development: lean canvas. Ash Maurya proposes an alternative version of the business model canvas: Lean Canvas. I find it simpler to use, and a slightly better fit for customer development. What I really like about it is that it has explicit fields for Problem (top 3 customer problems) and Solution (top 3 features of your product). Focusing on your customer, the problem and the solution is a good way to initiate your research. This is what Ash has to say about documenting hypotheses. He also provides an online lean canvas tool, allowing you to share your lean canvas with your partners or mentors.

Go ahead, fill one in. Start from the business model canvas you filled in before. I found that I tend to use the business model canvas as a (quick) brainstorm tool to think up alternative business models. Once I get serious about testing one of them, I will dive in deeper and fill all the fields of a lean canvas.

Lean Canvas, no-nonsense alternative way to document your hypotheses

What now? What you now have is a list of hypotheses about your business. If any of the things on your canvas mismatch reality, there is a chance you don’t have a viable business (yet). So you see why it’s so important to validate the hypotheses.

And you don’t have the answers – your customers do. So you need to get out of the building and get the answers from them.

In the next posts, I’ll write on how to do that.


  1. In scientific endeavours the method is to make an hypothesis and then perform experiments to see if the hypothesis can be refuted by the results of the experiment. In other words you look to see if there is proof of something that is not consistent with your hypothesis. Generally, findings that are consistent with your hypothesis are helpful but not conclusive as there may be other reasons that return this supporting evidence. Ultimately it comes down to sampling or how good of a predictor your sample size is of the whole. I guess in business model hypothesis testing what you are really doing is seeing if your assumptions have any basis in reality. So, the hypothesis would be for example: There are people who find that “x” is also a problem. An extension would then be: Do enough people find “x” as a problem. The latter is more difficult because the sample size comes into play. Maybe the 5 people you interview agree with you that “x” is a problem because they have read you blog posts on the problem. The first is really a test to see that your thinking is clear and the latter is about market size.

  2. Scott,

    thanks for clarifying the scientific approach. The interview approach you describe definitely does not rely on the statistical relevance, and is really about qualitative learning. The earlier you are in the process, the more uncertainty there is, and the coarser the experiments are.

    One thing that comes closer to scientific experimentation is split testing on large scale – provided you have large enough user base. You expose your “control group” – e.g. half of users – to your product as is, without any change, and the rest of the users to the version of your product with a new feature you are putting under test. You define upfront your hypotheses: e.g. this email notification feature will increase repeat use of the system by 10%. And then you measure repeat use of the system for the two groups and compare.

    In theory, if you see no change, you should actually remove the new feature, as it doesn’t create the assumed effect (and it does add to complexity of your product, both for the users, and for your developers).

  3. Here is some criticism of @ashmaurya ‘s lean canvas by Osterwalder himself (on Steve Blank’s blog)

  4. But some guys, they’ll do whatever they can to sell what they market as generic farmacia on line and Cialis that were coming from China. Orderingprescriptionsfrom an Farmacia On Line is legitimate or not? 50, if they want to have. When taken with alpha-blocking agents, for example, sent an email Thursday afternoon that didn’t actually
    depart his phone until 5 AM the following morning I went through his

  5. Hello, I think your site might be having browser compatibility issues.

    When I look at your blog in Firefox, it looks fine but when opening
    in Internet Explorer, it has some overlapping. I just wanted to give you
    a quick heads up! Other then that, excellent blog!

  6. On several occasions, she even admitted that she thought about boiled
    sweets mould as part of advertising or marketing campaigns with targeted messages,
    these items can routinely be purchased in copious quantities.

    Stop by my page: list old fashioned sweets uk

  7. Sugar-free sour candy will
    help in decreasing the threat of heart illnesses iin the diabetics.

    Before filling up the basket with chocolate-based goodies, double-check if
    the pewople receiving it prefer dark chocolate,
    and aim to eat only a small quantity of sweet syrup as he sucks.
    Investigators are looking into printing on coffee mugs and keyrings.

    Don’t waste your time. They will be delighted at these items.

  8. What’s Taking place i am new to this, I stumbled upon this I
    have discovered It positively helpful and it has
    helped me out loads. I am hoping to give a contribution
    & aid other customers like its helped me. Good job.

  9. Whats up are using WordPress for your blog platform?
    I’m new to the blog world but I’m trying to get started and create my own.
    Do you require any coding knowledge to make your own blog?
    Any help would be really appreciated!

    My web blog news 15; Maryann,

Leave a comment