Glossary

Explore the language of modern business intelligence and data-driven work. With plain-language explanations, practical context, and real-world examples, this glossary makes complex ideas clearer. From decision-making frameworks to the data stack behind AI and analytics, each entry is designed to spark smarter conversations.

Metrics Layer

A metrics layer is a centralized abstraction layer that sits between your data warehouse and your downstream analytics tools. It allows data teams to define business logic and KPI calculations—such as "Gross Margin" or "Monthly Active Users"—in a single, governed location. By decoupling metric definitions from individual dashboards, a metrics layer ensures that any tool or user querying the data receives the same, standardized result, effectively acting as the "semantic" source of truth for the modern data stack.

Data Analytics

Data Analytics is the practice of examining, transforming, and modelling data to uncover insights, answer questions, and support decision‑making. It turns raw records into signals you can act on, like patterns, trends, and relationships.

Dashboard

A Dashboard, in the context of data, is an interactive, visual interface that brings your key metrics into one place so you can monitor and act fast. It turns raw data into charts, tables, and cues that make change and performance easy to spot.

Ontology

Ontology, in the context of data and metrics, is the shared vocabulary that defines your business entities, metrics, relationships, and rules. It gives every term a single, trusted meaning across dashboards, queries, and AI powered analytics.

Metadata

Metadata means data about data. It describes a file, table, or metric so you can find it, understand it, and use it correctly. A photo’s metadata can include date, location, and camera model. A book’s metadata lists the title, author, and publisher.

Semantic Layer

A semantic layer is the shared business vocabulary and rules that translate raw tables into consistent, human‑readable metrics and dimensions. It turns questions like “What do we mean by revenue?” into reusable definitions every chart and query uses.

Data Integrity

Data integrity means your data stays accurate, consistent, complete, valid, and timely from the moment it is created to the moment you use it. Think of it like a shared recipe that everyone follows, so the result tastes the same every time.

Data Quality

Data quality measures the reliability of your data. High‑quality data is accurate, complete, timely, consistent across systems, standard-conformant, and free of duplication.

Headless BI

Headless BI is an approach to business intelligence where the metric and semantic layer sit behind an API, separate from built‑in visualizations. You define metrics once, then query those definitions from any front end, so every destination shows the same numbers.

Extract, Transform, Load (ETL)

ETL is a three‑step data process that helps you turn raw inputs into trustworthy information you can use. You extract data from multiple sources, transform it by cleaning and structuring it, then load it into a destination such as a data warehouse or lakehouse where your team can access it. Put simply, ETL is how you turn scattered, messy data into something clear and usable.

Extract, Load & Transform (ELT)

ELT (Extract, Load, Transform) is a modern data integration architecture that moves raw data from source systems directly into a target destination—typically a cloud data warehouse or data lake—before any processing occurs. Unlike traditional methods that require an external staging area, ELT leverages the massive computational power of the destination system to perform transformations. This "load-first" approach is preferred for handling large-scale, unstructured, or high-velocity data.

Metric Tree

A metric tree is a visual or conceptual model that maps how key business metrics relate to each other. It links a top‑level outcome, like revenue or retention, to the contributing drivers that explain changes underneath. You get a clear, shared view of cause and effect across teams.

Metric

A metric, in the context of analytics, is a calculated value that tracks performance for a business activity. Think of it as a consistent math formula applied to your data over time, like revenue, conversion rate, or churn rate. A metric includes a clear formula, time frame, and rules for how to slice the data. It turns raw numbers into a repeatable signal you can compare across periods, products, regions, or segments.

Business Intelligence

Business intelligence (BI) is the practice of collecting, organizing and analyzing data to help organizations make informed decisions. BI tools turn raw data into actionable insights through visualizations, reports and dashboards.

API

An API is a contract that defines how to request data or actions from a system, and what will be returned. Think of it like a restaurant menu and order slip. You ask for a dish, the kitchen prepares it, and you get exactly what you asked for.

Data Lake

A data lake is a centralized repository designed to store vast amounts of raw data in its native format—including structured, semi-structured, and unstructured data. Unlike a traditional warehouse, a data lake acts as a flexible "sandbox," allowing organizations to ingest data immediately and determine its schema only when it's ready for analysis.

Data Lineage

Data lineage maps the journey of your data from origin to destination. It visually shows where data comes from, how it’s transformed, and where it’s used.

Data Stack

A data stack is a set of tools, services, and procedures that work together to collect, process, store, and analyze an organization’s data.

Dimension

A dimension, in the context of data, is a descriptive attribute that provides context for your metrics. Think of dimensions as the categories or labels—like date, region, or product line—that you use to group, filter, or slice your data.

Data Catalog

A data catalog is an organized inventory of a company’s data assets. This centralized, access-controlled library typically lists datasets, tables, and fields alongside owners, definitions, and lineage so people can search, understand, and use data with confidence.

Cardinality

Cardinality describes how unique the values in a column are. It also plays a role in defining how tables relate to each other. A high-cardinality column contains many unique values, while a low-cardinality column contains few unique values.

Data Visualization

Data visualization is the representation of data as charts, diagrams, pictures, or tables. When information is presented visually, it’s easier to see patterns and quickly spot outliers. Data visualizations are also perfect for performing comparison and forecast analyses.

Data Warehouse

A data warehouse is a specialized, centralized repository designed to store, organize, and filter structured data from across an organization. Unlike operational databases that handle day-to-day transactions, a warehouse is architected specifically for OLAP (Online Analytical Processing). It provides a "single source of truth" for historical data, enabling businesses to perform complex queries and generate high-level business intelligence.

Key Performance Indicator (KPI)

A key performance indicator (KPI) is a measurable value that shows how effectively your organization is achieving its most important objectives. Think of KPIs like the gauges on your car dashboard—each one gives you real-time feedback to help you maintain your engine and stay on course.

Member

A member, in the context of data, is a specific, unique value within a dimension that represents an individual entity, category, or attribute. Think of a member as an item in a list—like “Q1 2025” in a list of time dimensions or “Blue T-Shirt” in a list of product dimensions.

Measure

A measure, in the context of data, is a quantifiable numeric value used to track and analyze data. It represents a calculation—like sum, average or count—that’s performed on raw data points.

Knowledge Graph

A knowledge graph is a structured network that represents real-world entities (people, places, products, metrics) and the relationships between them. It adds context to data, so systems and people can make smarter decisions.

MCP Server

An MCP server is an open‑standard service that exposes tools and resources to AI clients using the Model Context Protocol. It lets AI models retrieve live data and perform actions—securely and consistently—against systems like file stores, APIs, and development platforms.

Data Governance

Data governance is a formal framework of people, policies, and technology designed to ensure that an organization’s data assets are accurate, secure, and usable. Think of it as the "Librarian" of a massive digital library: every piece of data is cataloged, protected, and accessible only to those with the right permissions. In a business context, it establishes the rules for data stewardship, ensuring that information remains a reliable asset for analytics and stays compliant with privacy regulations.

Metric Catalog

A metric catalog is a centralized, governed repository of standardized business metrics and KPIs. It serves as an authoritative reference guide, documenting the precise name, calculation formula, and business context for every metric. By housing these definitions in a single location, a metric catalog eliminates "metric drift," ensuring that all departments—from Finance to Sales—calculate and interpret organizational progress using the exact same logic.

Objectives and Key Results (OKRs)

Objectives and Key Results (OKRs) are a goal-setting framework that helps teams align on ambitious goals and measure progress with specific, quantifiable results. Objectives define the destination on the roadmap, while 2–5 Key Results act as milestones that track success. Objectives inspire and guide teams, and Key Results keep everyone accountable by focusing on measurable outcomes.

Online Analytical Processing (OLAP)

Online analytical processing (OLAP) is a technology that makes it fast and easy to analyze large amounts of data from multiple angles. It organizes information into structures called "cubes" — think of these as pre-built summaries of your data — so you can explore and compare figures by time, location, product, or any other dimension, almost instantly.