Business of Machine Learning
The role of Machine Learning (ML) in business is an interesting topic. ML provides a powerful tool set for understanding data, but that is not enough. A fundamental tenet of strategy is differentiation, or finding a sustainable competitive advantage. We shall try to clarify basic alternatives below.
A technological answer to this situation is add context to data. One can employ a mathematical model reflecting deep domain knowledge to understand a data set better. Alternatively, one can combine multiple data sets to obtain information which would not have been possible with a single data set. It is expected that there is a lot of room for innovation in this area, given the increasing number of open data sets. An example of using multiple data sets is the MegaSense project of University of Helsinki and Bell Labs, in which abundant measurement data from cheap sensors is calibrated with a small number of high-quality sensors.
Finally, a powerful business concept may provide necessary differentiation to competition. Even though machine learning is by definition driven by data, ultimately it is the business concept which brings the cows home. Examples of a company building a dominating ecosystem before competition can be found in business literature, e.g. iTunes. The potential in this area is virtually unlimited, but ideally makes use of technical possibilities.
Suunto dive computers use mathematical model for nitrogen saturation of human body. For deep dives or challenging conditions, using multiple computers provides safety.
Exclusive access to data
This is the most talked about approach at the moment: a company has exclusive access to data and performs value adding operations in them. In the simplest case, the data in question are related to normal operations of the company, and may be Customer Relationship Management (CRM) data or enterprise business data, for example. Machine learning can be used to automate business processes using data.
A second type of exclusive data relates to a service product the company is offering. It may be based on two-sided network which links two domains, utilizing domain knowledge of the company. Examples of this approach include PayPal (linking businesses to customers) and Facebook (link businesses to information about consumers). Here, machine learning can be used to improve effectiveness, but it may also be central to the service concept. Facebook employs machine learning to select items to customers' newsfeed in an attempt to keep them loyal to the service, while serving the interests of its paying customers (advertisers).
For the first type, machine learning is an enabler for a strategic internal capability (increased effectiveness). Provided that the company is operating in an competitive environment, it is expected that each participant increases its effectiveness gradually. Thus, any competitive advantage in this area is typically not permanent.
For the second type, the impact of machine learning depends the extent there is relevant competition. There is no credible alternative to Facebook due to network effects and therefore its customers would not easily jump ship even if an alternative provider with improved functionality would emerge.
A second type of exclusive data relates to a service product the company is offering. It may be based on two-sided network which links two domains, utilizing domain knowledge of the company. Examples of this approach include PayPal (linking businesses to customers) and Facebook (link businesses to information about consumers). Here, machine learning can be used to improve effectiveness, but it may also be central to the service concept. Facebook employs machine learning to select items to customers' newsfeed in an attempt to keep them loyal to the service, while serving the interests of its paying customers (advertisers).
For the first type, machine learning is an enabler for a strategic internal capability (increased effectiveness). Provided that the company is operating in an competitive environment, it is expected that each participant increases its effectiveness gradually. Thus, any competitive advantage in this area is typically not permanent.
For the second type, the impact of machine learning depends the extent there is relevant competition. There is no credible alternative to Facebook due to network effects and therefore its customers would not easily jump ship even if an alternative provider with improved functionality would emerge.
Do clever things with data
If access to data is not a differentiating factor, there is no alternative but to use data in a more intelligent manner. The source of differentiation here is extracting more value from a data set than the competition. Application of an advanced algorithm to a data set may provide a short-term advantage, until the competition finds an even better substitute.A technological answer to this situation is add context to data. One can employ a mathematical model reflecting deep domain knowledge to understand a data set better. Alternatively, one can combine multiple data sets to obtain information which would not have been possible with a single data set. It is expected that there is a lot of room for innovation in this area, given the increasing number of open data sets. An example of using multiple data sets is the MegaSense project of University of Helsinki and Bell Labs, in which abundant measurement data from cheap sensors is calibrated with a small number of high-quality sensors.
Finally, a powerful business concept may provide necessary differentiation to competition. Even though machine learning is by definition driven by data, ultimately it is the business concept which brings the cows home. Examples of a company building a dominating ecosystem before competition can be found in business literature, e.g. iTunes. The potential in this area is virtually unlimited, but ideally makes use of technical possibilities.
Comments
Post a Comment