A crucial change in the operating environment, enabling the success of platform companies that utilise AI, has been the continuous increase in the number of various sensors and sources of data.
Thanks to the positioning of platform companies between users, providing a basis for their actions, these companies have primary access to users’ data. During the past five years, in particular, the increasing number of data sources has enabled platform companies to continuously improve their analytics capabilities. A contributing factor has been the development of machine learning. Thanks to better analytics capabilities, the platform can improve its offering through things such as customisation and better customer experiences. This way, it attracts new users and is able to offer new data sources, such as smart toasters, car heaters and barber appointments, to existing users.
More and more often, the pillars of platform economy are 1) data, 2) machine learning, and 3) digital platform.
The winning circle of platform companies
A business model cannot be protected through a digital platform. It is possible to find dozens of identical clones of the AirBnB website, for example. Rocket Internet in Germany has even turned the cloning of growing American platform companies into a successful business.
The protection of machine learning capabilities is mostly not possible, either. Repeating a breakthrough is considerably easier than making one. That is why companies such as Google, Facebook and Baidu typically publish their breakthroughs in machine learning in scientific articles. This is inconvenient for many smaller companies focusing on the development of AI and machine learning, because protecting their business in this way is their lifeline.
Therefore, data plays a key role in the protection of business. Data is so valuable that Andrew Ng, former chief scientist at Baidu, has said that large companies currently launch devices collecting data at prices lower than their production costs. One such product is Amazon Echo, a voice-guided speaker developed by Amazon.
Thus, the protection of business is based on the winning circle of platform companies, in which the advanced AI achieved through better data makes it possible to offer the digital platform to an ever larger number of people to an increasing extent. That is why some data intensive companies, such as Amazon, Baidu, Google, Tencent and Facebook, are currently invincible.
How can a Finnish company ever compete against these giants?
The value of data can decrease
The value of data can decrease in the near future. This possible decrease is due to two reasons. First of all, a new breakthrough, semi-supervised learning, is looming in the field of machine learning. It can be used to make more analyses with reasonable amounts of data resources with as reliable results as are produced with enormous masses of data. At the same time, the pressure to release data from public administration, on one hand, and the pressure to provide people with better control over their data, on the other hand, keep increasing. The EU is a forerunner in these matters. The PSD2 directive revolutionising the banking sector is a part of this development, as is the GDPR, which has given rise to a lot of discussion. From the point of view of EU member states, this is a positive development, because there are only a few platform companies with roots in Europe.
If the value of data goes down, what will the protection of business be based on? Some indication can be found by examining what is valuable from consumers’ point of view. A prevalent thought used to be that to the consumer, value is generated by the sources of data (1), such as activity bracelets. Next, there was talk about the value actually being in the data (2). It was valuable to know how many steps one had taken. Now people have come to the realisation that the data itself is actually worthless; it is the analysis (3) that provides the value, i.e. what kinds of health effects various numbers of steps provide.
However, that is not true, either. It is the insights (4) based on these analyses that are useful: ‘If you walk more, you will be healthier.’ Even this is not really enough. The actual value comes only when the user of the bracelet actually starts walking more. Actions (5) are valuable. And furthermore, health – the value promised through the bracelet – is only achieved when these actions become a pleasant routine (6) that promote a healthy lifestyle.
Traditional companies sell goods, i.e. their understanding of value is on level 1. The large platform companies are now on level 3. In other words, we have still a long way to go before we understand value. When value is understood more deeply, we can start thinking about new kinds of business models. What if activity bracelets were sold on the basis of verified health effects? Perhaps they could be paid for by the state or an insurance company, instead of the user.
Johannes Koponen works at Demos Helsinki, an independent think tank. His core competences lie in strategic foresight and the development of business models through scenarios and other methods used in futures studies. At Demos Helsinki, he has been in charge of extensive futures studies related to things such as work, health, and hyper-connected society. Koponen’s achievements include the founding of a start-up called Scoopinion, which won the Helsingin Sanomat Foundation’s media innovation contest, Uutisraivaaja, in 2011. He has also created a device for hospitals that turns children’s hearing test into a game.