How a startup can test hypotheses from a business case and not screw up

We tell you what a business case is and how you can use it to check the viability of a startup.

How a startup can test hypotheses from a business case and not screw up


Evgeny Kiryanov, head of the early-stage corporate gas pedal and a person who is a member of the expert committee in startup competitions and "Startup as a Diploma" programs at TPU, SPbSU, and SPbPU universities, is here.

I didn't put that in there to make a good point or to put some cool regalia on myself. It means that every year I review a huge number of early-stage and sometimes late-stage startups - literally thousands of startups every year. In 95% of them, I keep coming across one mistake that's easy to fix with Business Model Canvas. I will tell you what this mistake is and how to work with this model.

Startupers' main mistake

The mistake most startups make is to make unsubstantiated theses, or the complete absence of them where they are critically needed.

Every time a startup declares at a pitch, "We'll make a million because our invention - waterproof pita bread for shawarma - costs 100 rubles and we can sell it to 10,000 housewives through targeted VK advertising," a large number of theses must be formulated and proven:

  • housewives really make shawarma;
  • housewives do have a problem with soggy pita;
  • housewives don't know how to solve the pita problem;
  • housewives do sit on VK;
  • There are really more than 10,000 housewives on VK;
  • housewives are willing to pay 100 rubles for pita bread, etc.

Believe me, if you describe your product in answers to these and similar questions, and then intelligently formulate them as a pitch presentation, then investors and commission chairs will at least thank you.

Where to get a comprehensive list of hypothesis questions and how to work with it

Working with hypotheses is easiest through a kanban board, because the tasks for testing them are universal, not as big individually, and you don't have to involve most of the team responsible for business development, marketing, and product development. Since you have a startup, you do it yourself or with a cofounder.

The list of questions and subsequent hypotheses should be taken from the business case, otherwise known as the Osterwalder matrix. It sounds kind of scientific, but it's really quite simple.

You need to take a matrix of 9 blocks and draw arrows from one to the other and then read the text from the squares as the arrow moves, putting the question "Is it true?" at the beginning.

So, I quickly filled out the business case (totally bad, incomplete, embarrassing myself) at a level to make an example clear:

business cantvas sample

Next, draw arrows between the squares. For example, let's connect "Housewives" and "VK" as a promotion channel. Like this:

business case study

So far it's easy? Great!

Now let's start asking the question, "Is it true?" + "Housewives" + "VK". A bit of speculation magic and we get a good question: "Are housewives really on VK?" or "Is it true that housewives use VK?"

Answering such a question is quite realistic. Which means - we open a kanban board and task our cofounder to check "Is it true that housewives use VK?" To test the thesis, we need to filter users by employment and see how many of them are housewives or not working, but over 20 years old.

Formed a task with hypothesis testing in Shtab

In this way, we connect everything that connects. For example, everything to the right of the unique sentence connects to each other and to the bottom two blocks; everything to the left of the value sentence connects to each other and to the bottom two blocks. The value proposition connects to anything. But you'll figure it out for yourself when you start formulating questions.

After you have checked the simple questions, draw the arrows as double arrows. For example, continue our arrow to expenses.

We got "Indeed" + "Housewives" + "VK" + "10 rubles for the attracted client. Here the wording is more complicated, but let's try: "Are housewives really sitting in VK, and attracting customers costs 10 rubles?

Let's rephrase this one more time, changing the order for beauty, and we get the following: "Can we really attract housewives from VK for 10 rubles?

We add such a task to our kanban board. It's easy to check it: rivet an ad, run it for 500 rubles, and see how much it will cost to click and convert into a purchase.

Simple? Yes!

Does anyone do this before pitching? No.


It is very difficult to formulate all the important questions without visualizing them. Checking hypotheses without formulating them correctly is even more difficult. So take note - visualize, formulate, and only then check. And for convenience, write everything down in your project management system. For example, in Shtab.