The Business of Next Generation Technologies
In the limitless world of IoT and IR4.0, how do we identify effective business models?
THE FORECASTER’S ADVANTAGE
Consider a company that is developing a new product in the environmental monitoring space. The technology behind the device allows for the monitoring of a multitude of environmental conditions, including humidity, air quality (e.g., pollution), temperature, noise and light but also the presence of individuals and whether they move, they fall and so on. These characteristics can be activated or deactivated at the “touch of a button” from the sensor board.
In this situation, the challenge is not how to add a new characteristic, but to understand how best to help the user make informed decisions. This means different things for different people. You can have different environmental monitors for households, for offices, targeted to elderly and so on. This customer focus will help your company gain a competitive advantage by differentiating a good product from a bad one through better customer-centricity, even when they use the same technology.
You are daunted with the number of decisions you have to make. 1) What are you going to activate in terms of features given that you can easily add and subtract them? 2) Who do you want to target as a user given that the same technology can be packaged in a different box and software and made useful to any segment? 3) Which segment is going to pay for it and is that a subscription model, a one off or a combination of the two?
For example, in the case of the environmental monitor for households, the final payer is perhaps not the user but his or her family - or even an insurance company interested in predicting health deterioration in elderly patients that live at home.
The approach: An innovative adaptation of an effective approach, reference class forecasting, based on the work of Nobel Prize Daniel Kahneman
The situation above is a real one - and we sought to answer the question.
To do this, we developed an algorithm which looks into the “optimal” combination of characteristics to develop by considering 1) target users, 2) what competitors are doing in terms of already developed characteristics and 3) how successful competitors have been (for example, by studying the amount of funding raised by innovative startups).
The algorithm is based on Reference Class Forecasting, which is a method developed by Nobel Prize Daniel Kahneman and Amos Tversky for making predictions about the future based on similar cases from the past. Widely used in policy and planning, this is the first application to business modelling and marketing.
We created a unique dataset of startups and scale ups operating in the IoT in a specific industrial application: environmental monitoring. Our unique approach was comprised of three steps.
First, we started from publicly available data regarding startups, their founders and the amount of money they raised from Crunchbase, a well-known database. The overall database includes information from over 54,000 companies. From these companies, we filtered by fields and identified 9 broad markets which were related to environmental monitoring. We then randomly selected 149 startups that 1) were currently active, 2) had not been acquired, 3) had disclosed information about funding. From these, we narrowed down to 15 companies which were diversified in terms of their size.
Second, the company all had several use cases available on their websites. We collected a total of 200 use cases. We converted each case to a text file and then transferred it to a spreadsheet so it could be analysed using statistical software.
Third, we created a new algorithm to analyse this information. The algorithm is based on classifying the features of the use cases (both existing features and features that were missing) and comparing them with the particular sectors and the characteristics of the offering companies as well as the company at hand.
The value proposition generated is going to be both distinct from others in the market but also feasible by the company at hand given its capabilities. In the end, we identified that advantage would have been achieved by focusing on the provision of analytics as well as the security aspects of the solution proposed. These were the guiding principles for the identification of the features to develop straight away and the features that could be postponed.
The benefit/solution: a tight set of value propositions linked to the company’s capabilities and market prospect
When a company launches a new product in the marketplace, it typically investigates three main factors: (1) What their competitors are producing and which product characteristics need to be developed in order to be at least at par with their products, (2) Which product characteristics should be excluded because, for example, too expensive or technically difficult to make, and (3) Differentiating characteristics to develop in order to distinguish one product from others. The developed algorithm looks at all three. A company has already benefited from our algorithm as it uses the information obtained online about its competitors and their customers to identify critical characteristics it should be looking at developing.
In our study, the company at hand has ultimately identified 10 critical competitors, as well as 200 customers and 230 characteristics of their products. The analysis allowed the company to ultimately implement three types of decisions:
1) the sector in which sector to position itself
2) the particular features of the solution to develop, and ultimately
3) the type of customers to target.
One strength of our approach is that it uses information that is either publicly available on the internet or your company’s proprietary data collected over the years. This unbiased approach at developing business models and value propositions does allow you to filter through an enormous amount of data and help you prioritise what you need to do now to compete from what you can be postponed when the model has been tested in the marketplace.