Metric Verse
1. Overview
What is Metric Verse?
Metric Verse is a centralized system that organizes, connects, and manages key data components. It serves as the foundation for building reliable, reusable, and transparent metrics across the organization, ensuring consistency and alignment in reporting and analysis.
In Metric Verse, each metric is treated as a core asset — clearly defined, structured, and connected to related data elements through logical relationships. This approach helps eliminate discrepancies by creating a single source of truth for all metrics across the workspace.
Purpose and benefits of Metric Verse
The primary purpose of Metric Verse is to achieve clarity, consistency, and control over how data is measured, reported, and analyzed. By providing a standardized structure for metrics and their dependencies, it helps reduce errors, improve collaboration, and speed up the process of generating meaningful insights.
Key Benefits:
Consistency: Everyone in the organization works with the same definitions and logic for key metrics.
Transparency: Clear connections between metrics, dimensions, and measures make reporting and analysis more understandable and auditable.
Efficiency: Reusable components save time when building reports, dashboards, and analyses.
Governance: Better control over how data is defined, filtered, and used within the business.
Scalability: Supports growing datasets and evolving business questions without compromising data integrity.
One of the key advantages of Metric Verse is that it evolves with your activity. As you run queries, verify data, or make changes to components, the system automatically reconciles the platform. AI tracks all dependencies, cascades updates, and adjusts metrics accordingly. There's no need for manual refreshes or chasing down disconnected components.
Where to find it
You can access Metric Verse directly from the Context Engine section in the main navigation menu of the platform.
To locate it, follow these steps:
Click on the Context Engine icon on the left-hand navigation bar
In the dropdown menu, select Metricverse.
This will open the Metric Verse interface where you can manage and explore all components.
2. Metric Verse components
Metric Verse is not a collection of isolated components — it is a fully connected system. Every metric, subject, filter, and dimension is linked together through an AI-driven hierarchy, from high-level subjects down to base-level columns. This unified lineage is maintained without requiring manual SQL materialization. Each metric is a living object that reflects real-time dependencies and is updated with the data model - something that’s nearly impossible to maintain manually at scale.
2.1 Metrics
Metrics are the foundation of the Metricverse. They represent the core business measurements that organizations use to track performance, monitor progress, and make informed decisions. Examples of metrics: "Average Discount", "Average Order Value", "Average Fulfillment Time".
Metric structure
Name: A clear, descriptive label (e.g., Average Discount for European Suppliers)
Description: Explanation of what the metric measures
Expression: The formula or logic used to calculate the metric
State: Indicates the metric's current status:
Generated: Automatically created by the system.
Verified: Reviewed and validated by a user
Unverified: Marked as unverified by a user
Verify: A column showing if the metric has been reviewed.
On click
When you click on a specific metric in Metricverse, you are presented with a detailed breakdown of how that metric is structured and how it works within your data model. The screen includes the following elements:
Description: A concise explanation of what the metric measures and its business purpose.
Expression: The formula or logic used to calculate the metric (e.g., Average Discount, Total Price, etc.).
Referenced Entities:
Dimensions: Data categories involved in the calculation (e.g., Supplier Name).
Measures: Any pre-defined calculations the metric uses (e.g., Average Discount).
Filters: Any applied conditions that scope or limit the data (e.g., High-Value Order Filter for orders > $10,000).
Referenced By: Shows user questions or queries where this metric is actively used.
Visual Diagram: A visual representation of how the metric is connected to dimensions, measures, and filters.
You can click on any element within the visual diagram to drill deeper into its details. This makes it easier to navigate through Metric Verse and understand how each component contributes to your analytics.
2.2 Subjects
Subjects identify the business topic or area of focus — such as "Customer Spending" or "Sales Performance". A subject serves as a logical grouping of data components that are related to a common business context. This helps teams quickly locate the data assets relevant to their area of interest.
Subject structure
Name: A clear, descriptive title identifying the business topic (e.g., Customers, Orders, Line Items).
Description: A brief explanation of what the subject covers (e.g., "Customer records with RFM analysis").
Key: The primary field used to uniquely identify records within the subject (e.g., Customer Key, Order Key, Line Item Number).
State: Indicates whether the subject is ready for use (e.g., Generated).
On click
When you click on a subject, Metric Verse provides a detailed view that shows:
Description of the subject (what it analyzes)
Primary key field (used for linking data)
Referenced Entities — columns that belong to the subject (e.g., Customer ID, Order Date, Order Priority)
Referenced By — specific user questions or queries in the system that utilize this subject
Visual diagram showing how attributes, dimensions, filters, measures, and custom calculations are connected within the subject
2.3 Measures
Measures represent quantitative calculations in Metricverse. They are typically numerical values aggregated from the underlying data using functions such as SUM, AVG, MIN, MAX, or COUNT. Measures serve as the building blocks for constructing metrics, offering consistent and reusable calculations across the dataset.
Measure structure
Name: A clear, descriptive label (e.g., Average Delivery Time, Active Status).
Description: A brief explanation of what the measure calculates.
Expression: The aggregation logic or formula used (e.g., AVG(Receipt Date - Ship Date)).
Subject Name: Identifies the data subject area the measure belongs to, such as Orders, Line Items, or Customers.
Type: Indicates whether it's a custom measure or an aggregation measure.
State: The current status of the measure:
Generated: Automatically created by the system
Verified: Reviewed and validated by a user
Unverified: Marked as unverified by a user
Verify: A column showing if the measure has been reviewed.
On click
When you click on a measure, you will see detailed information structured as follows:
Name: The name of the measure (e.g., Average Delivery Time).
Description: Explanation of what the measure calculates.
Type: Whether the measure is an Aggregate Measure or another type.
Expression: The specific fields or formulas used to compute the measure (e.g., receipt date minus ship date).
Aggregator: The aggregation logic applied (e.g., Average, Sum).
Referenced Entities: Which columns (fields) from the data are used in this measure.
Referenced By: Indicates:
The user question that this measure helps to answer.
The metrics that rely on this measure.
Just like with metrics, clicking on different items allows you to drill down for further details and see dependencies.
2.4 Dimensions
Dimensions are data categories used to group, filter, and segment metrics within the Metric Verse. They help structure your data in a meaningful way, enabling you to analyze performance across various categories. Examples of dimensions include Customer Nation, Product Type, or Region. They are essential for breaking down and interpreting metrics in a structured and relevant manner.
Dimension structure
Name: A clear, descriptive label (e.g., Customer Key, Customer Country).
Description: Explains what the dimension represents.
Expression: Shows the logic or SQL used to calculate the dimension.
Subject Name: Specifies which subject (e.g., Orders, Customers) this dimension belongs to.
Type: Specifies the type of data the dimension represents:
Categorical: For grouping data by discrete categories (e.g., Country, Customer ID).
Temporal: For working with dates or time periods (e.g., Order Date, Ship Date).
Ordinal: Indicates whether the dimension has a meaningful sort order (Yes/No).
Sort Key: Displays the key used to sort values within the dimension, if applicable.
State: Indicates the dimension's current status:
Generated: Automatically created by the system.
Verified: Reviewed and validated by a user
Unverified: Marked as unverified by a user
Verify: A column showing if the dimension has been reviewed.
On click
When you click on a dimension, you will see:
The dimension's name, description, type, and expression.
The Ordinal field indicating whether the order of values matters.
The Referenced Entities section showing columns from the data.
The Referenced By section, which shows:
Which user questions this dimension supports.
Which metrics depend on this dimension.
2.5 Attributes
Attributes are raw data fields within a subject, providing descriptive context about the entities represented in your dataset. They help you identify, label, and provide context for the metrics and dimensions you analyze.
Attribute structure
Name: A descriptive label (e.g., Customer Name, Customer ID).
Description: A short explanation of what the attribute represents.
Expression: The formula or column used to derive the attribute.
Subject Name: Identifies the subject (e.g., Orders, Customers) to which this attribute belongs.
State: Indicates the current status of the attribute:
Generated — created by the system.
Verified: Reviewed and validated by a user
Unverified: Marked as unverified by a user
Verify: A column showing if the attribute has been reviewed.
On click
When you click on an attribute, you will see:
Name and Type: The name and the label "Attribute."
Description: What the attribute represents.
Expression: SQL expression or column name.
Referenced Entities: Data columns linked to this attribute.
Referenced By:
User questions that reference this attribute.
Metrics connected to this attribute.
2.6 Filters
Filters narrow down the data within a subject based on defined conditions — for example, specific time periods, regions, or customer segments. They help refine and focus the dataset used for analysis, making it easier to zoom in on high-performing groups or isolate data for targeted insights.
Filter structure
Name: A clear, descriptive label for the filter (e.g., Active Order Filter (6 Months)).
Description: A brief explanation of the condition the filter applies (e.g., Checks if the customer's last order date is within the last 6 months).
Expression: The SQL condition or logic used to define the filter (e.g., Customer Region = 'AFRICA').
Subject Name: The subject the filter is associated with (e.g., Orders or Customers).
State: The current status of the filter:
Generated: Created by the system.
Verified: Reviewed and validated by a user
Unverified: Marked as unverified by a user
Verify: A column showing if the filter has been reviewed.
On click
When you click on a filter, you’ll see detailed information organized as follows:
Name: The name of the filter.
Description: A brief explanation of the filter’s purpose.
Expression: The specific SQL logic the filter uses.
Referenced Entities: Which columns the filter relies on (e.g.,
c_nationkey).Referenced By: Which questions and metrics this filter supports.
2.7 Relationships
Relationships establish how different elements within the Metricverse (such as metrics, measures, dimensions, attributes, and filters) are connected to each other. These connections allow data to flow between components, supporting complex queries and analytical operations. Relationships typically connect data sources through specific keys, forming a data model backbone.
Relationship structure
Name: A descriptive label that defines the connection (e.g., Customer to Nation).
Description: Clarifies the nature of the relationship between datasets (e.g., Each customer belongs to a single nation).
Kind: Specifies the relationship type (Direct, Lookup, etc.).
Cardinality: Describes the data connection structure (e.g., Many-to-One, One-to-Many).
State: The current status of the relationship:
Generated: Created by the system.
Verified: Reviewed and validated by a user
Unverified: Marked as unverified by a user
Required: Specifies whether this relationship is mandatory for data integrity.
Unique on To: States if the relationship is unique on the destination side.
Max Matches: Shows the maximum allowed matching records.
Multiplicative: Indicates whether the relationship causes data duplication through multiplication.
Distinct: Shows if the relationship enforces distinctness.
Verify: Shows if the relationship has been reviewed.
On click
Clicking on a relationship opens a detailed view containing:
Type: Direct, Lookup, etc.
Cardinality: many_to_one, one_to_many, etc.
Required: Yes/No
Unique on To: Yes/No
Max Matches: Maximum allowed matching records
Multiplicative: Yes/No
Distinct: Yes/No
From Table / To Table: The source and destination entities involved in the relationship.
Keys: The columns that form the join keys.
Referenced Entities: The columns from each table used in this relationship.
Referenced By: Lists queries where this relationship is being used.
3. Best Practices for Using Metric Verse
To ensure you get the most value from Metric Verse, follow these best practices when building, maintaining, and analyzing your data components:
Use Subject Pages to Explore Business Topics
Start with a subject (e.g., Orders, Customer Activity) to understand the metrics and logic grouped around a specific business area. This gives you context before diving into individual components.
Follow the Metric-to-Column Lineage
Use the visual diagrams to trace a metric’s logic down to the exact measures, filters, relationships, and even table columns it relies on. This helps you understand how a number was produced and why it’s trustworthy.
Click Through Components to See What Drives Results
Metrics in Metric Verse are interactive. Click into dimensions, filters, or measures from any metric diagram to inspect the logic behind each part. It’s your way of “auditing” the reasoning without reading SQL.
Look at the Referenced By Section to Understand Impact
Each component shows where it is used — in metrics or user questions. This helps you understand how widely it is relied on and whether it plays a critical role in business reporting.
Rely on Verification Status to Identify Trusted Logic
Use the verification badges to prioritize components that have been reviewed and approved for business use. If a metric or filter is unverified, treat it as a candidate for further scrutiny.
Use Filters to Decode Business Logic
Filters often represent strategic business definitions (e.g., “High-Value Orders”). Reviewing filters helps you understand how your organization segments and interprets data.
Trust the AI for Structure — Focus on Interpretation
You don’t need to manage joins, modularity, or deduplication. Instead, spend your time making sure the logic makes sense from a business point of view.
Treat Metricverse as a Living Source of Truth
Because the system updates itself when queries run or terms change, you’re always looking at the most current, connected, and consistent logic. Use it as your authoritative reference — not dashboards or slide decks.
Use Metricverse to Investigate Inconsistencies
If numbers don’t match across tools, trace them in Metricverse. You’ll likely uncover whether filters, measures, or relationships differ — and can pinpoint the root cause with full visibility.
Ask Better Questions Using What You See
Use the structure and logic in Metricverse to improve how you formulate business questions. You’ll start asking things like “What metric already includes shipping delays?” or “Can I reuse this filter to answer X?”
Last updated