Lotta Hassi

Two years ago we started to work closely with several companies to learn how these organizations could become more experimentation-driven in their innovation activities. This means developing ideas for e.g. new offerings or processes through a series of fast, small, low-cost experiments.One of the things we are paying close attention to in these explorative projects is the emotional experience of being engaged in an experimentation-driven process.

Moving from a conceptual idea to action and realizing parts of the idea in order to learn more about it is difficult. At that initial point, you only have a rough idea. Placing that idea in front of a customer to receive feedback is a leap into the unknown. The uncertainty of these moments makes the development team feel vulnerable and the fear of failure makes it more attractive to retreat into the perceived safety of planning because “the idea is not yet ready”. The ability to tolerate incompleteness and uncertainty and to move forward regardless of these feelings requires great effort. Interestingly, at the same time, doing early experiments provides a boost to the energy level of the team. The team’s initial energy is directed towards making the idea come alive through small experiments, and the feedback from the experiments becomes a self-feeding loop, fueling the team’s interest towards learning more, making the idea become increasingly real – and refined.

In short, the experimentation-driven approach starts with an idea you want to develop, identifying the key assumptions regarding that idea, and coming up with creative ways to test those assumptions in practice. Key assumptions are those that can either make or break the idea. In the case of Zappos, which has grown to become the largest online shoe store, the key assumption in the beginning was “are customers willing to buy shoes online, without seeing and trying them on before purchase?” Where a business school educated MBA graduate would have approached this topic by spending hour after hour doing market research, crafting a business plan – which in itself is a fictitious document describing a reality that is definitely not going to happen – and carefully calculating cash flow potential, the founder of Zappos did something else: He went to a local shoe store, asked the owner if he could photograph the shoes and put those for sale online. If someone would make a purchase, he would then buy the shoes himself from the store and ship them to the customer. No IT systems, no warehouses, no marketing. Just a rudimentary website that helped validate the key assumption, without which there would be no business. And in the process he also learned a thing or two about what kind of payment options to offer, how to handle returns, how to do customer service and so on. (2, 3)

The basic assumption behind experimentation-driven innovation is that uncertainties cannot be resolved by planning. As a method it shifts the focus to learning as much as possible about your idea, quickly, while keeping costs low. Furthermore, when experimentation is the main tool, what you learn will be based on experience, not on assumptions. In other words, you get actual proof about the validity of your idea. This is especially true when your idea has anything to do with human behaviour as opposed to something that is purely mechanistic. For example, in one of the cases I studied a simple idea aimed to increase motivation and decision-making of employees lead also to increases in productivity, teamwork, trust between workers and the supervisors, and perhaps most surprisingly got the workers to proactively start taking ownership of and improving the internal work processes in the team. It might be too far off to say that none of these effects could not have been seen beforehand by thorough planning, but even so, without actually experimenting the idea there would have been little proof about them.

In fact, quite often it is simply less costly and much faster to do a hands-on experiment to see what happens instead of planning all the possible scenarios inside one’s head. As my boss in the MIND research group has a habit of saying; “show, don’t tell.”

  Perhaps the most interesting thing that has happened in the field of entrepreneurship research in the past 15 years is the work done by Saras D Sarasvathy from University of Virginia and her colleagues. She studied how expert entrepreneurs – meaning people who have built businesses with annual sales between $200 million and $6.5 billion – approach starting a new venture. And guess what? They pretty much use approaches similar to the experimentation-driven method, focusing on what they can do with their existing means, using low-cost ways to validate their key assumptions, and getting the first customers and other stakeholders on-board. In other words, they focus on acting instead of planning, control instead of prediction, and without risking what they cannot afford to lose. (4, 5)

It is becoming something of a cliche to say that the world is getting more and more complex, and that uncertainty is similarly increasing as a result. However, the advent of Big Data has not made us much better at uncovering uncertainties, especially in emerging contexts where there might not exist any data. This calls for a model of rational decision-making that is not rooted in the Newtonian worldview of clear cause-and-effect relationships and linear thinking. When it comes to creating something innovative, the focus needs to shift from emphasis on prediction to emphasis on control – meaning that you put your energy into what you can do, here and now, to develop your ideas and keeping options open so when uncertainties inevitably do happen, you are able to learn from them and adapt your approach.

The future is constantly being created and shaped by human action. It does not exist ‘out there’ to be predicted. And this means it is up to you to help create the kind of future you want.

You start with an initial opportunity idea for a new business. At this point you have an idea of an opportunity, but don’t know how to best realize it, or even, if it can ever be successfully realized. To test the opportunity idea, you need to create experimentation ideas, that is, how to test the opportunity idea with the target customers with minimum resources invested. What can we do in order to learn what we should do? Then, from these experiments you collect execution ideas – lessons learned on what works and what doesn’t, how the opportunity idea should be executed in order to build a successful business.

Your opportunity idea can be, a grocery home-delivery service for sub-urban areas – a way to ease the everyday life of families with kids. To test the value proposition, you search for the first five families to buy into your idea. Your experimentation idea is to deliver personally the service for a month to the first five families that are willing to take part in your experiment. You find an apartment building, where five families decide to pay you 10€ per week, they send their shopping list to you by email, you find the best store to do the shopping for them, collect their groceries, pack them in your car, and bring them to the agreed place in their home, which can be a simple fridge in a garage, to which all the families have access. From this experiment, you collect the execution ideas: how to organize for the collecting of the shopping lists, how to find the best deals for the items from the near-by supermarkets, how to optimize the pick-up and delivery of the groceries, and so on. And above all, you find out whether you have a market to begin with.