Business Model Components

A business model describes how an organization creates, delivers, and captures value.

Machine Learning is enabling new business models that can out compete more traditional models, resulting in increased:

Growth: new business models can create access to new customers using new products and services
Profitability: the successful implementation of key business model components can lead to high profit margins
Barriers to Competition: organizations that harness new business models can grow to leading positions that are difficult to challenge

Examples of high profile companies that have successfully implemented new business models with significant Machine Learning components include:

These new models leverage four key components:

User Networks: leverage the knowledge and interaction of product and service users
Machine Learning: leverage the power of computing and algorithms
Human Learning: leverage the adaptability and wisdom of human intelligence
Data: leverage the vast and growing repository of data

The components and their interconnections are shown in the graph below:

The change over time produced by incremental improvements in the key components and their effects on other components is significant. Integrated over time, the aggregate change can be expressed as:

$\\A = \int_{0}^{t} \sum_{i=1}^{n}\left ( \triangledown C_{i}+\sum_{j=1}^{n}\left ( \triangledown C_{i}C_{j\neq i} \right ) \right ) \\\\\mathrm{\textup{where:}} \\\\A\,\,\,\,\,\mathrm{\textup{aggregate change}} \\\triangledown \,\,\,\,\,\mathrm{\textup{change in a component}} \\C\,\,\,\,\,\mathrm{\textup{component (User Network, Machine Learning, Human Learning, Data)}} \\t\,\,\,\,\,\,\,\mathrm{\textup{time}} \\n\,\,\,\,\,\,\mathrm{\textup{number of components}} \\i\,\,\,\,\,\,\,\,\mathrm{\textup{index over components}} \\j\,\,\,\,\,\,\,\mathrm{\textup{index over components}}$

User Networks

User networks are the fabric of connections and uses created as the networks grow. This is often referred to as the Network Effect. Related to the Network Effect is Metcalfe’s Law which states that the effect of a telecommunications network is proportional to the square of the number of connected users of the system.

Examples of User Networks contributing significantly to the success of new business models include:

Google: user search patterns provide data for advertising
Netflix: user viewing data provides data for movie recommendations
WhatsApp: more users provide more opportunity for messaging
Facebook: more users attract ever more users
Amazon: more users provide more reviews and recommendations

Machine Learning

Machine Learning contributes to the success of new business models for a number of reasons:

Perception: ML models are able to process extremely fine grain inputs, such as for image recognition
Memory: trained ML models don’t forget; models can be retrained to learn from new data
Speed: trained ML models are able to process predictions in seconds or fractions of a second
Accuracy: many ML models have achieved accuracies above those of humans
Cost Effectiveness: ML models can be duplicated and deployed world-wide at low costs relative to the benefits provided
Growth: the power of ML is doubling every 3.4 months

Human Learning

Human Learning is a critical component in new business models for adaptation to ever more rapid changes in:

Technology: technological change is advancing rapidly in many areas, including:
- Machine Learning: e.g. algorithms, applications
- Cloud: e.g., Google, Microsoft, Amazon
- Smartphones: e.g., Google, Apple, Samsung
- Apps: e.g., for smartphones, desktop/laptop
- Social Media: e.g. Facebook, Twitter, WhatsApp, Instagram
- Computing: e.g., faster CPUs, graphics processing units
- Telecommunications: e.g., LTE, 4G, 5G, Internet
Competition: new business models are unleashing new competitive forces
Organization: new business models are necessitating new organizational structures

Data

Data is a critical component of new business models due to data:

Growth in Data: the cumulative year over year growth in volume of data is estimated at over 60%
Machine Learning: data is needed for ML model training and prediction processing
Probability and Statistical Analysis: to understand ML performance and business conditions

References

Wikipedia Article