Develop InnovationsTools

Hypotheses & Experiments

When we develop an innovative business model we try to paint a picture of the future. Naturally some aspects of this picture are well understood and we know we can manage or implement them successfully. Other aspects of the picture are unclear or uncertain, we don’t understand them well and we don’t have evidence if they will be possible at all. To deal with those uncertainties we usually define assumptions and hope that they will be true. Customers will behave in a certain way, the cost of customer acquisition will be below a specific value, and we can realize a function of our product with medium implementation efforts, to name a few. Many businesses fail because those assumptions were wrong or overstated and they only find out after a lot of time and money has been wasted.

In Business Design we know that we don’t know many things. In order to minimize the risk and reduce the impact of those assumptions on our business we have defined a process for hypotheses development based on analogs and antilogs. We can test those hypotheses and create evidence based certainty that our assumptions are true or false before spending lots of money and time on them.

A hypothesis is often defined as an educated guess because it is informed by what you already know about a topic. This step in the process is to identify all hypotheses that merit detailed examination, keeping in mind that there is a distinction between the hypotheses generation and hypotheses evaluation.

If the analysis does not begin with the correct hypotheses, it is unlikely to get the correct answer. Psychological research into how people go about generating hypotheses shows that people are actually rather poor at thinking of all the possibilities. Therefore, at the hypotheses generation stage, it is wise to bring together a group of analysts with different backgrounds and perspectives for a brainstorming session. Ideally you use a template like ours to assist with the visualization of the hypotheses and experiments.

You can use our template as a printout DIN A0 poster and fill it with sticky notes or use our online version in Rapid Modeler.

  1. Analogs: There are aspects in every innovative business model that are new to you, your team or your company. Maybe you want to address a new customer segment with your product or services. You know as well that some competitors already offer similar products or services to exactly that customer segment. When Apple developed the IPOD they knew that Sony already served many consumers with mobile music through the Walkman. So others have already proven that the market for mobile music exists and that people pay for devices. This knowledge helps to reduce the risk when you plan to behave similar to the competition.
  2. Antilogs: Most likely your new innovative business model shows elements or a combination of them that are new for you, your team, your company and even the industry. You will not be able to learn from somebody else and to avoid costly assumptions you have to find ways to prove and validate them before implementing your new business. Apple didn’t know if people are willing to pay for a downloaded song. They knew that people download digital media but looking at the commercial failure of Napster, SoundJam or LimeWire they had no evidence that anybody is willing to pay 99 cents for a song.

In reality it is not always easy to define what an analog or antilog is. If a competitor does something you want to copy, it is not certain that your organisation will be able to this successfully. Often little details like different branding perception of the customers, internal processes, structures in your partner channel, power play within your value chain can make a big difference. An “analog” suddenly becomes an “antilog”.

To reduce effort and increase effectiveness we define two dimensions within the found antilogs as the next step. One dimension looks at the importance of an antilog to our business model. Will our model collapse if an antilog doesn’t work? The other dimension looks at the uncertainty of an antilog. Our test and experiment focus now concentrates on those antilogs that are both important and uncertain.

In our validation exercise we define two questions:

  1. What hypotheses grow out of relevant antilogs in our test focus?
    “We believe……” –> “….which will result in ….”
  2. How can we test the identified hypotheses with the least effort?
    “We do….” –> “We are convinced, if….” (incl. threshold and timeframe)

In most business models it is sufficient to concentrate on the top 3-5 antilogs for testing. As Business Design is an iterative process your results will most likely alter elements of your business model or minimal viable product / business.