Mindware Research Institute offers a quantitative business strategy management methodology using self-organizing maps to existing consulting firms, market research firms, data science service companies, and freelance consultants and data scientists. This methodology has been used as a top-secret method by some of the world’s leading companies and consulting firms over the past 20 years. Mindware Soken is currently building a community (Mindware Salon) to train evangelists while compiling these experiences.
Traditional strategic management methods such as SWOT analysis and PPM (Product Portfolio Management) are now outdated. This is because in traditional methods, strategies are expressed in qualitative information (natural language), and once formulated, they are fixed and cannot respond to rapid changes in the external environment.
One of the principles of scientific management that Tada learned at the Japan Management Association in the past is that “what cannot be quantified cannot be controlled. This means that in conventional strategic management, strategy cannot be subject to control. Strategies contain complex information and cannot be expressed in a single number. To make a strategy controllable, it is necessary to build a mathematical model that includes the various parameters associated with the strategy. When I say this, not only businessmen but also traditional strategy consultants will be frightened. However, modern technology makes such a thing possible.
With SOM (Self-Organizing Maps), it is possible to create a literal strategy map from numerical data. The user does not need to pay much attention to mathematical problems. That is the role of those who program computer software. However, the user must be exploratory enough to avoid drawing false conclusions. It is similar to the drawing technique in that it “pauses the use of existing concepts and sees things as they are.
The basis of the strategy is to avoid dispersing your company’s position as much as possible (or to eliminate the enemy from your position). If we look at the map from the point of view of “which areas” will strengthen our position, the answer is obvious: we define them as target areas (battlegrounds) in the SOM, and we know the value of each attribute (variable) in those areas. The software we use has a powerful profile analysis feature that tells us which attributes are statistically significantly higher or lower than other areas. This allows us to make quantitative decisions about what to increase (or decrease) and by how much to increase (or decrease) in order to gain that area.
Although positioning maps using principal component analysis have existed in the past, it may be easier to say that they are, roughly speaking, a more powerful version of such maps. Whereas the positioning map represented each brand (competitor) only as a point on a coordinate plane, the SOM provides richer information.
For example, a map based on consumer survey data orders consumers by similarity in terms of demographic attributes, needs, lifestyles, hobbies, preferences, uses, and involvement (with each brand), etc. Each hexagonal piece on the SOM is called a “node” Each node contains the most similar consumers, and adjacent nodes are more similar to each other than to their neighbors. Thus, the SOM identifies similar consumers, e.g., those who are highly involved in each brand, not in points, but in a spread. The characteristics of each node or set of nodes can then be evaluated using statistical methods with each attribute of the data.
In addition to creating a map of the entire data set, you can select or weight only the attributes of interest for more refined analysis. For example, to see the relationship between needs and level of involvement, a map can be created by applying appropriate weights to multiple needs attributes and leaving the level of involvement attribute with zero weight to visualize the existence of a relationship. Attribute (variable) selection and attribute weighting implies adjusting the perspective of the analysis. It also allows us to relate attributes that were not included in the original data. For example, when selecting target areas, the attribute “profitability” would be more obvious. Other examples include the linkage of needs and seeds, and the association of consumer product preferences in sensory evaluation analysis (preference mapping) with the characterization of each product by experts in the field.
By joining our community, you not only receive a software license for use in quantitative business strategy management, but you also receive ongoing support through our hotline support (Slack or email). We intend to gradually enhance the range of presentation materials and proposal templates for member consultants to provide to their clients, as well as handbook-like materials on the methodology and videos for self-study.
Mindware Salon Membership Information
Future Market Structure Analysis