A
TERM:
Agent
CATEGORY:
AI Packaging Pattern
An AI agent is a software program designed to autonomously perform specific tasks or solve problems by using artificial intelligence. These agents can interact with users, systems, or environments to achieve predefined objectives. They often rely on machine learning and natural language processing techniques to adapt and improve their performance over time.
TERM:
Agentic AI
CATEGORY:
AI Foundation
Agentic AI refers to artificial intelligence systems capable of autonomous action and decision-making. These systems, often called AI agents, can pursue goals independently, make decisions, handle complex situations, and adapt to changing environments without direct human intervention. They leverage advanced techniques such as reinforcement learning and evolutionary algorithms to optimize their behaviour and achieve specific objectives set by their human creators.
TERM:
AI-as-a-Service (AlaaS)
CATEGORY:
AI Packaging Pattern
A cloud-based service model that provides artificial intelligence capabilities to businesses without the need for extensive in-house AI development. This allows companies to leverage AI technologies through APIs or pre-built models.
TERM:
AI Bias
CATEGORY:
AI Foundation
AI Bias refers to systematic errors in AI systems that can lead to unfair or discriminatory outcomes. These biases can stem from training data, algorithm design, or human prejudices inadvertently incorporated into the AI system. AI used in pricing models must be carefully tested for bias.
TERM:
Ambient AI
CATEGORY:
AI Foundation
AI that is available anytime, anywhere, often through devices.
TERM:
Application Credits
CATEGORY:
Pricing Methodology
Fungible credits can be used to pay for a variety of different functions and are transferable from one function or product to another.
TERM:
Agent Family
CATEGORY:
AI Packaging Pattern
A collection of Agents performing related functions or tasks. Agents in a family can often be sequenced to perform a more elaborate business process or in some cases substituted for each other.
TERM:
Artificial Intelligence (AI)
CATEGORY:
Discipline
Artificial intelligence (AI) is a branch of computer science that aims to build machines capable of performing tasks that typically require human intelligence. AI enables machines to simulate human abilities, such as learning, problem-solving, decision-making, and comprehension.
C
TERM:
CasualML
CATEGORY:
AI Foundation
CausalML is a subfield of machine learning that focuses on understanding and estimating causal relationships between variables.
TERM:
Category Value Map
CATEGORY:
Customer Value Management
A Customer Value Map is a visual tool used to analyze and communicate how a product or service creates value for customers. It typically consists of two main components:
The Value Map: This describes the products and services offered, how they create customer gains, and how they alleviate customer pains.
The Customer Profile: This outlines the customer's jobs to be done, their pains, and their desired gains.
The Customer Value Map helps businesses align their value propositions with customer needs, enabling them to identify areas for improvement and differentiation from competitors.
TERM:
Chain of Thought
CATEGORY:
Prompting Strategy
Chain of Thought (CoT) is an AI reasoning technique that enhances the problem-solving capabilities of large language models by breaking down complex tasks into smaller, logical steps.
TERM:
Compute Costs
CATEGORY:
AI Cost Factor
Compute costs refer to the financial expenses associated with the computational resources required for artificial intelligence systems to perform tasks such as processing data, training machine learning models, and making predictions. These resources can include hardware like GPUs and TPUs, as well as cloud computing services.
TERM:
Context Based Pricing
CATEGORY:
Pricing Methodology
Pricing methodology proposed by Mark Stiving where the business context is used to determine the best approach to packaging and pricing.
TERM:
Co-Pilot
CATEGORY:
AI Packaging Pattern
An AI Co-Pilot is a smart assistant powered by artificial intelligence that collaborates with users to boost productivity, decision-making, and creativity through real-time suggestions, automation, and insights.
TERM:
Computational Intelligence
CATEGORY:
AI Foundation
An alternative to the transformer based approach used by Wolfram Alpha to compute answers to questions. Especially powerful in math related applications.
TERM:
Consumption Pricing
CATEGORY:
Pricing Methodology
An approach to pricing in which the pricing metric is the amount of a service actually.
TERM:
Cost Plus
CATEGORY:
Pricing Methodology
Cost Plus is a pricing methodology where a fixed percentage or amount of profit is added to the cost of a product or service. This approach ensures that all costs are covered and a profit is made on every sale.
D
TERM:
Deep Learning (DL)
CATEGORY:
AI Foundation
Deep learning is a subset of machine learning that uses artificial neural networks with multiple layers to learn from large amounts of data. It enables computers to process information in ways similar to the human brain, allowing them to recognize complex patterns and perform tasks such as image recognition, natural language processing, and autonomous decision-making.
TERM:
Design Structure Matrix (DSM)
CATEGORY:
Customer Value Management
A Design Structure Matrix is a square matrix used to represent and analyze the structure of complex systems or processes. Key features of a DSM include:
It lists system elements along both rows and columns in the same order.
Off-diagonal cells indicate relationships or dependencies between elements.
It can represent various types of relationships, such as information flow, material flow, or energy transfer.
DSMs are particularly useful in systems engineering and project management for visualizing dependencies, identifying feedback loops, and optimizing system structures. The attached document uses DSMs to analyze value drivers in the Learning Solutions category, demonstrating their applicability in business strategy and product development contexts.
TERM:
Diffusion Model
CATEGORY:
AI Foundation
Diffusion models are a type of generative AI model that progressively adds random noise to data and then reverses the process to generate high-quality outputs. They are widely used in applications such as image generation, video synthesis, and sound design and are known for their ability to produce highly accurate and detailed results.
TERM:
Disruptive Innovation
CATEGORY:
Innovation Approach
Disruptive Innovation, a term coined by Clayton M. Christensen, refers to a process where a smaller company with fewer resources successfully challenges established incumbent businesses by introducing innovations that disrupt the market. These innovations often start at the bottom of a market and relentlessly move up, eventually displacing established competitors.
TERM:
Distillation
CATEGORY:
AI Foundation
A process for transferring a bigger model’s capabilities to a smaller one.
TERM:
Dynamic Pricing
CATEGORY:
Pricing Methodology
An AI-driven strategy that allows prices to fluctuate based on real-time factors such as market demand, customer behavior, and competitor pricing. This approach optimizes pricing for profitability and competitiveness.
E
TERM:
Economic Value
CATEGORY:
Customer Value Management
Economic value is the worth of a good or service determined by people’s preferences and the trade-offs they choose given their scarce resources.
TERM:
Elastic Access Model
CATEGORY:
Pricing Model
A flexible monetization approach that allows for dynamic adjustment of prices and packaging, enabling businesses to adapt quickly to market changes and customer needs.
TERM:
Embedding
CATEGORY:
AI Foundation
Embedding is a means of representing objects like text, images, and audio as points in a continuous vector space where the locations of those points in space are semantically meaningful to machine learning (ML) algorithms. Embedding enables machine learning models to find similar objects and learn complex patterns and relationships in the data.
TERM:
Explainable AI (xAI)
CATEGORY:
AI Foundation
A set of processes and methods that allows human users to comprehend and trust the results and output created by machine learning algorithms. It aims to make AI systems more transparent and interpretable, addressing the "black box" problem often associated with complex AI models. Not to be confused with the foundation model developer XAI.
F
TERM:
Federated Learning
CATEGORY:
AI Foundation
Federated Learning is a machine learning technique that enables training models on distributed datasets without centralizing the data. It allows multiple parties to collaboratively train a model while keeping their data locally, enhancing privacy and data security.
TERM:
Few-Shot Learning
CATEGORY:
AI Foundation
Few-Shot Learning is an AI approach where models are trained to recognize new classes or perform new tasks with very few examples, typically ranging from 1 to 5 samples per class. This technique is particularly useful in scenarios where large labeled datasets are not available Relative to B2B cases where there tend to be relatively small data sets.
TERM:
Foundation Model
CATEGORY:
AI Foundation
A large language model, generally with more than one trillion parameters, that power many AI applications.
TERM:
Frontier Model
CATEGORY:
AI Foundation
A model that is at the leading edge of development based on its size (number of parameters) or on some performance criteria.
TERM:
Fungible Token or Credit
CATEGORY:
Pricing Methodology
A fungible token is a type of digital asset that is designed to be identical in value and interchangeable with other tokens of the same type.
G
TERM:
Generated Value Model
CATEGORY:
Customer Value Management
A value model generated using generative AI by a value model generation process (VMG).
TERM:
Generative Pricing
CATEGORY:
Pricing Methodology
Generative Pricing uses generative AI to create new pricing models and strategies. It leverages AI’s ability to analyze vast amounts of data to develop innovative pricing approaches, including dynamic pricing, personalized pricing, and value-based pricing. This methodology aims to create pricing models that are more responsive to market conditions, customer behaviour, and competitive dynamics, ultimately leading to more effective and profitable pricing strategies. It is based on a concept blend of value based pricing.
TERM:
Generator
CATEGORY:
AI Packaging Pattern
A product or function using a model capable of producing new data examples that resemble the training data it was provided.
TERM:
Good Better Best (GBB)
CATEGORY:
Packaging Pattern
See Tiered Pricing (GBB).
TERM:
Growth Motion
CATEGORY:
Go-to-Market Tactic
A Go-to-Market tactic: Product Led Growth, Sales Led Growth, Service Led Growth, AI Led Growth, Community Led Growth, Partner Led Growth, Relationship Led Growth.
H
TERM:
Hybrid Pricing Models
CATEGORY:
Pricing Model
Pricing strategies that combine multiple approaches, such as blending subscription-based pricing with usage-based components to offer more flexibility and predictability for both customers and vendors.
I
TERM:
Input Token
CATEGORY:
AI Pricing
Input tokens are the pieces of text that you provide to a language model as input. These tokens can be individual characters, words, or sub-words, depending on the model’s tokenization process. They are used by the model to understand the meaning and context of the input and generate a response.
TERM:
Input:Output Ratio
CATEGORY:
AI Pricing
The ratio of the price of an input token to the price of an output token.
J
TERM:
Jevons Paradox
CATEGORY:
Innovation Approach
A term coined by 19th-century British economist William Stanley Jevons to describe the following observation: Even when we develop new technology that more efficiently uses some commodity—coal, in Jevons’ day—it may increase, not decrease, demand for that commodity.
TERM:
Job Based Pricing
CATEGORY:
Pricing Methodology
Pricing methodology proposed by Gary Bailey for pricing agents. Leverages Clayton Christensen Jobs-to-be-Done model.
K
TERM:
K-Means Clustering
CATEGORY:
AI Approach
K-Means Clustering is an unsupervised learning algorithm used to group unlabeled data into clusters based on similarity. It partitions the data into a predefined number of clusters (K) by minimizing the variance within each cluster. The algorithm iteratively assigns data points to the nearest cluster centroid and recalculates the centroids until convergence. It is widely used in applications such as market segmentation, image compression, and anomaly detection.
L
TERM:
Large Action Model (LAM)
CATEGORY:
AI Foundation
Designed to learn and execute complex sequences of actions, often used in robotics and autonomous systems.
TERM:
Large Audio Model (LAM)
CATEGORY:
AI Foundation
Focused on processing and generating audio data, including speech recognition and music generation.
TERM:
Large Knowledge Model (LKM)
CATEGORY:
AI Foundation
Focused on storing and retrieving vast amounts of factual information.
TERM:
Large Language Model (LLM)
CATEGORY:
AI Foundation
A Large Language Model (LLM) is a type of artificial intelligence model designed for natural language processing tasks such as language generation and understanding. These models utilize deep learning techniques and are trained on vast datasets to predict and generate text. They are foundational models in AI, often based on Transformer architectures, and are used in applications like chatbots, content creation, and sentiment analysis.
TERM:
Large Multimodal Model (LMM)
CATEGORY:
AI Foundation
These models can process and generate content across multiple modalities, such as text, images, and audio.
TERM:
Large Reasoning Model (LRM)
CATEGORY:
AI Foundation
These models emphasize logical reasoning and problem-solving capabilities, implementing approaches such as Chain of Thought or Self-Discover.
TERM:
Large Video Model (LVM)
CATEGORY:
AI Foundation
Designed to understand and generate video content, combining aspects of vision and temporal processing
TERM:
Large Vision Model (LVM)
CATEGORY:
AI Foundation
Specializing in processing and understanding visual information, these models are crucial for computer vision tasks.
TERM:
Large Simulation Model (LSM)
CATEGORY:
AI Foundation
Used for creating detailed simulations of complex systems, such as climate models or economic forecasts.
TERM:
Leiden Clustering
CATEGORY:
Innovation Approach
Leiden Clustering is a community detection algorithm used in network analysis. It identifies groups of nodes that are more densely connected to each other than to the rest of the network. The algorithm optimizes modularity, ensuring that the identified communities are well-defined and meaningful. It improves upon the Louvain algorithm by addressing issues such as poorly connected or disconnected communities.
M
TERM:
Menu Options
CATEGORY:
Pricing Methodology
A packaging pattern where the buyer selects from a set of options for each category, much like a restaurant menu.
TERM:
Mixture of Experts
CATEGORY:
AI Foundation
A model in which only some specialized parts (experts) are active at a time, making the model cheaper to use.
O
TERM:
Open Source Model (OSM)
CATEGORY:
Pricing Foundation
A machine learning system that grants users the freedom to use, study, modify, and share its components without restriction. This typically includes access to:
The model's weights and parameters
The source code used to train and run the model
Detailed information about the training data
Under the Open Source Initiative definition, truly open-source AI should allow users to freely inspect, adapt, and redistribute the system for any purpose without needing permission. This approach aims to foster transparency, collaboration, and innovation in AI development.
TERM:
Open Weights
CATEGORY:
Pricing Foundation
Open-weights AI models are large language models (LLMs) whose parameters or "weights" are publicly accessible and can be used or modified without restriction. These models allow users to download and run them in various contexts, subject to certain limitations and licenses. Open-weights models differ from closed or proprietary models by offering more transparency and control, enabling businesses and researchers to adjust the relationships among the model's parameters to suit specific applications.
TERM:
Outcome Based Pricing
CATEGORY:
Pricing Methodology
Outcome-based pricing (also known as results-based pricing) is a pricing model in which the cost of a product or service is based on the value or outcomes it delivers to the customer rather than a fixed agreed cost. Instead of pricing being directly linked to the cost of the product or service, it is tied to the performance or outcomes achieved.
TERM:
Output Token
CATEGORY:
AI Pricing
In the context of AI, particularly language models like GPT, output tokens are the units of text that the model generates as a response to the input tokens. These tokens are the building blocks that allow the model to interpret and generate language.
P
TERM:
Packaging
CATEGORY:
Packaging Pattern
Standard approaches to combining functions, data, and services so that they can be priced and sold.
TERM:
Parameters
CATEGORY:
AI Foundation
The billions of small numbers that make up an AI model and determine its behavior.
TERM:
Platform + Extensions
CATEGORY:
Packaging Pattern
Platform + Extensions refers to the integration of additional functionalities or capabilities into an existing platform, particularly in the context of AI. These extensions are designed to enhance the platform’s performance, expand its range of tasks, or adapt it to specific applications. For example, AI extensions can include new algorithms, APIs, or tools that enable real-time data processing, task automation, or other advanced functionalities.
TERM:
Pricing Metric
CATEGORY:
Pricing Model
The unit of consumption for which a buyer pays.
R
TERM:
Reasoning Model
CATEGORY:
AI Foundation
Reasoning models in AI are systems designed to perform logical inferences, problem-solving, and decision-making tasks. These models often employ techniques like chain-of-thought reasoning to break down complex problems into smaller, manageable steps, enabling them to synthesize information and provide contextually relevant responses.
TERM:
Reasoning Token
CATEGORY:
AI Pricing Metric
A token generated in OpenAI o1 and 03 models and used as a pricing metric.
TERM:
Reinforcement Learning From Human Feedback (RLHF)
CATEGORY:
AI Foundation
Reinforcement Learning from Human Feedback (RLHF) is a machine learning technique that combines reinforcement learning with human evaluation to train and optimize AI models, particularly large language models. It involves using human feedback to create a reward model that guides the AI's learning process, helping it generate outputs that are more aligned with human preferences and expectations
S
TERM:
Scaling Laws
CATEGORY:
AI Foundation
An observation that a model’s performance tends to reliably improve given more parameters, training data and computing.
TERM:
Self-Discover
CATEGORY:
Prompting Pattern
A prompting pattern where the LLM chooses the best reasoning structure to tackle complex reasoning problems. An alternative to the chain of thought.
TERM:
Service as Software
CATEGORY:
AI Packaging Pattern
The ‘Service as Software’ business model involves automating tasks previously performed by humans while maintaining the same user interface. The backend processes are replaced with algorithms or robotics, enabling digital or physical workflow automation.
TERM:
Sustaining Innovation
CATEGORY:
Innovation Approach
Sustaining Innovation refers to incremental improvements to existing products, services, or processes to maintain competitive advantage and meet customer needs. It focuses on enhancing existing offerings rather than creating entirely new markets or products.
TERM:
Synthetic Data
CATEGORY:
AI Foundation
Synthetic Data is artificially generated data that mimics real-world data. It is created using algorithms and is particularly useful for training AI models when real data is scarce, sensitive, or expensive to obtain. Use of synthetic data for pricing and monetization is an important area of research.
T
TERM:
Tiered (GBB)
CATEGORY:
Packaging Pattern
The Good-Better-Best (GBB) pricing strategy involves creating three price bands or tiers for different product or service bundles. Each level includes better features or functionality than the one below, encouraging consumers to upgrade to a better product. This strategy is based on the premise that customers consider aspects like quality and value, not just price when making decisions.
TERM:
Token
CATEGORY:
AI Pricing Metric
See Input Token, Output Token, and Reasoning Token.
TERM:
Token-based Pricing
CATEGORY:
Pricing Methodology
A monetization model where users purchase tokens or credits that can be exchanged to access various AI functionalities. This approach allows for flexible pricing and usage metering.
TERM:
Tokenization
CATEGORY:
AI Foundation
Tokenization is the process of breaking down data, such as text, into smaller units called tokens. These tokens are essential for machine learning models to process and understand the data, enabling tasks like natural language processing, computer vision, and audio analysis.
TERM:
Transformer
CATEGORY:
AI Foundation
“All you need is attention.” A Transformer is a type of neural network architecture that processes sequential data, such as natural language, by using mechanisms like self-attention to understand context and relationships within the data. It is widely used in natural language processing and other AI applications.
TERM:
Training Data
CATEGORY:
AI Foundation
The data (text, images, videos and so on) used to develop an AI model. AI developers use data as a chisel to sculpt the parameters of their models. The model learns from each piece of data until its performance improves—sometimes exceeding the skills of human experts. For LLMs, training data comes in two stages. In pre-training, the model learns to predict text from trillions of words gathered from the internet and other sources. In post-training, more refined training data—including that collected from RLHF—helps mold the pre-trained base model.
U
TERM:
Usage-based Pricing
CATEGORY:
Pricing Model
A pricing model where customers are charged based on their actual usage of the AI service, such as the number of API calls, amount of data processed, or computational resources consumed.
V
TERM:
Value
CATEGORY:
Pricing Foundation
There are three types of value: Economic Value, Emotional Value, and Community Value. All are relevant to pricing, but Economic Value plays the leading role.
TERM:
Value Based Pricing
CATEGORY:
Pricing Methodology
Value based pricing is a strategy of setting prices primarily based on a consumer’s perceived value of a product or service. Value based pricing is customer-focused, meaning companies base their pricing on how much the customer believes a product is worth.
TERM:
Value Capture Ratio
CATEGORY:
Customer Value Management
Price / Value to Customer. Also known as the Value Ratio. Used to target pricing levels in value based pricing.
TERM:
Value Cycle
CATEGORY:
Customer Value Management
The Value Cycle is a management framework that defines how organizations create, communicate, deliver, document, and capture value. It is an iterative process used to sustain superior performance by exceeding customer needs and aligning value creation with pricing strategies.
TERM:
Value Driver
CATEGORY:
Part of Value Model
A factor determining the value of a solution to a customer. There are six main types:
Increase Revenues
Reduce Operating Costs
Reduce Operating Capital
Reduce or Defer Capital Investment
Reduce Risk
Increase Options
TERM:
Value Driver Equation
CATEGORY:
Part of the Value Model
An equation used to quantify a value driver.
TERM:
Value Driver Variable
CATEGORY:
Part of the Value Model
A variable in a value driver equation, usually specific to a customer, is a configuration combination. There are four types of variables:
Solution Variable
Customer Variable
External Variable
Improvement Claim
TERM:
Value Metric
CATEGORY:
Part of the Value Model
The unit of consumption by which a buyer gets value.
TERM:
Value Model
CATEGORY:
Customer Value Management
A Value Model is a system of equations that estimates the economic value (dollar value) a solution provides to the buyer. It is foundational to value-based pricing and market segmentation, quantifying the impact of a solution on profit and loss statements and balance sheets. Value models are used to reason, explain, design, communicate, act, predict, and explore the economic impact of solutions.
TERM:
Value Model Generation (VMG)
CATEGORY:
Customer Value Management
The use of generative AI and related technologies to generate a value model and keep it updated. Generative AI can also be used to customize a value model for a specific customer.
TERM:
Value to Customer (V2C)
CATEGORY:
Customer Value Management
The value that a solution provides to a customer. It can be cumulative over the life of the solution or for a specific time interval.
W
TERM:
Willingness to Pay (WTB)
CATEGORY:
Pricing Methodology
Willingness to Pay (WTP) is the maximum price a customer is willing to pay for a product or service. It is typically represented as a dollar figure or price range and reflects the perceived value of the product or service.
