Glossary of Terms

A structured way to represent information about entities and their relationships, commonly used in AI systems.

Acronym for Large Language Model, referring to AI systems trained on vast text datasets.

The percentage of visitors who take a desired action on a specific landing page, like signing up or starting a trial, on a specific landing page.

An advanced AI model trained on vast text data to generate human-like language outputs.

The beginning of a marketing project to prepare the sales and marketing team to connect with customers and prospects. Not all launches are aligned with a development release.

A family of open-source large language models developed by Meta (formerly Facebook).

A mathematical formula that measures how far off an AI model's predictions are from the correct results.

An easy product or feature change that can substantially improve metrics. For instance, adding a button to the homepage for users to purchase a product might be a “low-hanging fruit.” (Note: Low-hanging fruit is picked by amateurs. Experienced harvesters choose from the top of the tree, not the bottom.)

The subset of AI focused on building systems that learn patterns from data rather than following explicit programming instructions.

Data formats structured for easy processing by computers.

A market (or marketplace) is the collection of all the buyers and sellers in the area or region under consideration. Some product professionals use market and market segment interchangeably.

A market segment is a group of people or organization who share one or more common characteristics, grouped together for marketing purposes. Some product professionals use market and market segment interchangeably.

The technique of asking an AI model how to create better prompts for itself or other AI systems.

Early stage of a product or service, with just enough features to share with early adopters and garner insights for validation and improvement. A better term is "Minimum Viable Prototype.")

“Native” often refers to apps running on a smartphone or computer. Web apps are those accessed via a URL. For instance, the Facebook app is a native app. Facebook.com on a mobile Google Chrome browser on one’s smartphone is on the mobile web platform.

A document that explains the details of an AI model, including its purpose, training data, limitations, and ethical considerations.

The learned parameters within an AI model that determine how it processes inputs and generates outputs.

A module is a single-function programming unit within a software product. For example, spell-checking is a module in most word processing software.

The number of unique users who engage with the product monthly.

The predictable revenue generated by a subscription-based product on a monthly basis.

AI models that process multiple types of media like text, images, and audio.

Involving or using multiple modes or types of communication or information processing.

Human language that AI systems are trained to understand.

A field of AI focused on enabling computers to understand, interpret, and respond to human language.

A measure of customer satisfaction and loyalty based on how likely users are to recommend the product to others.

A type of machine learning model inspired by the structure of the human brain, consisting of layers of interconnected nodes (neurons).

A prompt engineering technique that avoids biased language to get more balanced and objective AI responses.

The single metric that best captures the success of your product. Of course, no metric will tell a full story, but the north star metric is the most important metric to guide product decisions.

Objectives and Key Results is a goal-setting tool used to set challenging, ambitious goals with measurable results. OKRs are how you track progress, create alignment, and encourage engagement around measurable goals. (See also KPI).

An offering is a grouping of products and services packaged to address a market problem. In some cases, offering is used to describe a one-time proposal for a client, such as described in an RFP.

An open-source framework for running large language models locally on personal computers.

The percentage of users who complete the onboarding process successfully.

There are always more ideas than resources. Opportunity scoring is a technique for comparing multiple projects to one another to select the one (or few) that will move to the next step of business planning.

A condition where an AI model becomes too specialized on training data, reducing its ability to perform well on new data.

Drawing broad conclusions from limited evidence or specific instances.

A pain point is a problem that prospective customers or users are experiencing. Pain point identification is helpful to generate targeted solutions.

The process of defining or determining the parameters or variables that control a system or function.

An AI-powered search engine that generates concise answers.

An archetype of the typical buyer or user of a product. A persona definition usually includes a name and role, a description, and common problems they encounter.

A suite of products that are packaged for a specific type of buyer. For example, Microsoft 365 (ie., Microsoft Office) is a package containing software products such as Word, Excel, PowerPoint, and Outlook as well as services of Teams and OneDrive.

An early release of a product’s launch to validate the go-to-market efforts with customers and sometimes sales prospects. (Not to be confused with Beta Test).

Creating and interpreting new data using a method designed specifically to the research objective. The research instrument directly answered the hypothesis. Popular techniques include interviews, observation, surveys, experiments, and prototypes. (See also Secondary [Market] Research

Using techniques to determine which feature or project is the most important. A prioritization scheme usually incorporates some combination of value to client, value to the business, and effort to deliver. Popular prioritization techniques include Kano, RICE and BRICE, Weighted Small Job First (WSJF), and IDEA.

A form of story focused specifically on the problem. Often "user stories" are written about a feature, rather than the problem or requirement.

In practice, a product is a set of functional capabilities or features that address a specific problem.

The percentage of users who adopt new features or functionality within a certain time frame after release.

Product lifecycle refers to the stages a product goes through from its initial idea, development, introduction to the market, growth, maturity, and eventually, its decline and removal from the market. The product lifecycle is typically represented as a curve with the four stages: introduction, growth, maturity, and decline.
 
During the introduction stage, the product is launched and gains acceptance from early adopters. In the growth stage, the product gains widespread acceptance, and sales volume rapidly increases. In the maturity stage, sales growth slows as the market becomes saturated with competitors and product innovation declines. In the decline stage, sales decrease as the product becomes obsolete or is replaced by newer products.
 
Effective product management is critical to extend the lifecycle of a product and maximize profits.

Product managers represent the market and the business of the product. They identify market problems and customer friction that a product or feature will fulfill, articulate what success looks like for a product, and rally a team to turn that vision into a reality.

The product marketing manager (also called product growth manager) focuses on sales enablement, go-to-market readiness, and growth strategy so that when a product is delivered, people want to buy it.

As described in the Scrum Guide, a Scrum Product Owner is accountable for maximizing the value of the product resulting from the work of the Scrum Team. How this is done may vary widely across organizations and teams. There is considerable confusion in the industry about this role and its relationship to the product manager role.

Product strategy defines the vision and goals for a product, identifies market opportunities, prioritizes market problems to solve, validates the concept in the market, and ensures a net-positive outcome for the business.

As it relates to “Project,” a program is an on-going effort to implement and maintain a product or project. Building a bridge is a project; maintaining the bridge is a program.

Also known as “big room planning,” this agile method helps teams of all types and sizes align around goals, business objectives, and customer needs. Generally, it’s a quarterly gathering that brings everyone together — from software developers to stakeholders — to complete essential planning of a large project.

An individual or collaborative enterprise that is carefully planned to achieve a particular aim. A project typically has a limited duration and scope. Building a bridge is a project; maintaining the bridge is a program.

Project managers are responsible for planning, executing, and closing projects. They are accountable for ensuring the project meets its objectives, is completed on time, and stays within budget.

The practice of crafting effective prompts to improve the accuracy and relevance of AI-generated outputs.

An individual or organization actively engaged in a sales process.

A restructuring of the existing computer code. There are a variety of reasons for a code refactor. Often, refactors are advantageous because they accelerate product development or reduce vulnerabilities.

The percentage of referrals that convert into active users.

A type of machine learning where an agent learns by interacting with its environment and receiving rewards or penalties.

The end of a development project that results in a product that is ready for customers. Not all releases have formal launches.

A project plan containing anticipated deliverables and rough date estimates.

Also known as a Invitation to Tender (ITT), an RFP is request from a client to a number of vendors which lists their requirements to address a client problem.

A description of a problem that will be addressed by a product feature or service. Some requirements are called non-functional requirements that include address issues related security, compliance, performance, or user-interface.

The RACI matrix is a simple but effective model for defining, assigning, and documenting roles and responsibilities in a project or task. RACI stands for Responsible, Approves (sometimes ‘Accountable’), Consulted, and Informed.

A technique combining AI text generation with real-time retrieval of information for improved accuracy.

The ratio between net income (over a period) and the spending necessary to deliver some capability.

The amount of revenue lost due to customer cancellations, downgrades, or non-renewals within a specific time period.

Product sales revenue for a specific period such as monthly, quarterly, or annually. It’s surprising how many product managers do NOT have access to this information.

High-level summary of a product’s vision and direction over time. A roadmap is a visual prototype of your strategy with phases of work shown in a sequential pattern. Roadmaps should not contain features and dates.

A strategic planning and strategic management technique used to help a person or organization identify strengths, weaknesses, opportunities, and threats related to business competition or project planning. [from Wikipedia]

The process of ensuring the sales organizations, both direct and indirect, have the knowledge and skills to successfully sell the product or service. Sales Enablement teams work with the sales organization to understand their needs and ensure the required content and training is created with the help of subject matter experts throughout the organization, particularly product management and product marketing.

Scrum is a process framework used to manage product development and other knowledge work. Scrum provides a means for teams to establish a hypothesis of how they think something works, try it out, reflect on the experience, and make the appropriate adjustments.

A Scrum Master ensures a team follows agile principles and practices as described in the Scrum Guide. They act as facilitators and coaches for the team, helping to remove obstacles and improve team performance. A Scrum Master is not a project manager or product owner.

The Scrum Master is accountable for establishing Scrum as defined in the Scrum Guide. They do this by helping everyone understand Scrum theory and practice, both within the Scrum Team and the organization. The primary role is coach of the Scrum method, not as a project or people manager.

A Scrum Team is a collection of individuals working together to deliver the requested and committed product increments.

Infer the answer to your hypothesis by leveraging data gathered without your research objective. That is, using other people’s data to guide your decisions.

When an AI critiques and improves its own outputs.

The study of meaning in language or the relationship between words and their meanings.

The average number of sessions per user within a specific time frame.

The duration of a user’s interaction with the product during a single session.

A solution addresses a market problem. In a tech-enabled business, a solution is often a package of products and services designed to address a market problem. (See Portfolio).

Stories are short requirements or requests written from the perspective of an end user. Sometimes called “epics” or "user stories." Not to be confused with "Success Story" or "Customer Case Study."

The time for a request to go through the creation process, from idea to design, development, and deployment. Analyze the time stories spend in each step of your process. Delays can indicate chaos in the creation or build process, reveal that the product team is not adequately staffed to handle the requested work, or that the product manager is not rejecting enough requests.

A profile of a customer who has achieved success with the product. Also called a "Customer Case Study.." It usually explains the scenarios before and after successful implementation.

Artificially generated data used to train or test AI models when real data is limited or sensitive.

TAM, SAM and SOM are acronyms that represent different subsets of market demand. TAM (Total Available Market) is the total market demand for a product or service. SAM (Serviceable Available Market) is the portion of the TAM targeted by your products and services which is within your reach. SOM (Serviceable Obtainable Market) is the portion of SAM that you can capture. These values can be measured in count (that is, number of potential clients) or revenue (amount of potential purchases).

The percentage of users who successfully complete a specific task or set of tasks within your product. This metric helps assess the usability and effectiveness of your product’s features.

A classification system that organizes items into groups based on their similarities and differences.

In software development, technical debt (also known as design debt or code debt) is the implied cost of additional rework caused by choosing an easy (limited) solution now instead of using a better approach that would take longer. [from Wikipedia]

A parameter controlling the randomness in AI text generation, with higher values producing more creative but potentially less accurate outputs.

Organizational approaches to label backlog items, epics, and initiatives into logical groups of work. Themes should inspire the creation of epics and initiatives but don’t have a rigid 1-to-1 relationship with them. A theme for a rocket ship company might be “Safety First” while an initiative would be “Successfully launch into orbit.”

The time it takes for a user to experience the core benefits of your product after starting to use it. A shorter TTV leads to higher user satisfaction, engagement, and retention. In a product-led growth, optimizing TTV is crucial to ensure users quickly understand the value your product delivers.

The basic unit of text processed by AI language models, typically a word, part of a word, or a character.

The process of breaking text into smaller units, such as words or subwords, for processing by an AI model.

The breakdown of incoming user traffic by different sources, such as organic search, referrals, or paid ads.

Converting speech, music, or other audio content into written or printed form.

A method where a pre-trained model is adapted to new tasks to improve performance.

AI architectures using attention mechanisms to process sequential data, forming the foundation of most modern language models.

The percentage of trial users who convert into paying customers.