Data Warehouse Fundamentals: Data Dimensions and Measures

Businesses today are collecting massive amounts of data as they continue to aggressively roll out and advance digital transformation initiatives. 

At the same time, the vast amount of intelligence that organizations are collecting isn’t being put to use. In large part, this is because companies lack the ability to discover and act on it in a timely manner. Oftentimes, this is because data lives in multiple, disparate databases that lack interoperability.

This is a big problem, especially when considering the need for data—and, more specifically, event data—in today’s business environment. Simply put, companies that can’t quickly access intelligence and string together data points stand little chance of matching data-driven competitors.

Add it all up, and a gap is now forming between data-driven organizations and laggards. And it appears as though this trend is only going to worsen in the years to come. 

In an effort to become more data-driven, a growing number of organizations are implementing cutting-edge data warehousing strategies. Keep reading to learn about how data warehousing works; some fundamental aspects, including data dimensions and measures; and how Scalyr fits into the picture.

A data warehouse signifying data dimensions and measures

What Is Data Warehousing?

Data warehousing involves pulling together information from multiple sources so it can get analyzed and compared to drive business decisions. Data warehouses are subject-oriented and time-variant in nature.

Are Databases Data Warehouses? 

Databases and data warehouses are similar, but they serve different purposes within an enterprise IT environment. 

In short, databases store transactional data. The point of a database is to provide end users and systems with access to information as it flows in from different endpoints. Databases are relational in nature, with rows, columns, and rules. 

Data warehouses, on the other hand, connect with multiple databases and systems, bringing together large volumes of transactional and informational data and making it readily accessible in a centralized location. Think of it just as the name suggests: a warehouse filled with tons of data.

data dimensions and measures

What Are Facts and Measures in Data Warehouses?

Businesses run on various events called “facts.” Some examples of facts may include the total number of sales in a particular location, the number of customers who have joined a loyalty program, or the average rate of purchase for various products during a specific time of the year.

For example, a national retailer may want to analyze winter coat sales in Minnesota during the first three weeks of November. The organization may also choose to break down their data further by analyzing performance across different stores across the state or region. 

By analyzing facts, companies can get highly granular with their data and uncover critical growth trends. 

How Fact Tables and Measures Work

Whenever a fact occurs, it gets stored in a location called a fact table. This contains specific measures or descriptive figures about the event. Fact tables also work in conjunction with one or more lookup tables or master lists. 

For example, a measure might be the price of a particular transaction. This table may contain several columns with further descriptions—like the date of the sale, the location, and the customer ID.

Types of Fact Tables 

Now that you have a better idea of what fact tables are, it’s important to understand the various types of fact tables that businesses can use when conducting data warehousing. 

Additive Fact Table

Additive facts get summed up through the different dimensions in a fact table.

Non-additive Fact Table

Non-additive facts can’t get summed up for any dimension in the fact table.

Cumulative Fact Table

A cumulative fact table explains what happened over a specific period of time, such as total sales by product store per week. 

Snapshot Fact Table

Many enterprises are now using snapshot fact tables to analyze data at specific points in time.

What Are Dimensions?

Fact tables require the structure of dimensional tables. Dimensions are structures within a star schema that describe and categorize fact tables. 

To clarify, the main purpose of a dimension is to support a fact table and offer descriptions about what it contains. Dimensions live in dimension tables. Dimension tables can range from a handful of rows to millions of rows, depending on the size and scale of the operation. 

What Is the Star Schema?

For those who may be unfamiliar with the term, a star schema is a database structure that’s commonly used in data warehouse and business intelligence deployments. 

The star schema uses a single fact table to store measured and transactional data. It also has smaller dimensional tales that contain data attributes.

Just as the name suggests, a star schema resembles the shape of a star. 

Common Types of Dimension Tables 

Time Dimension 

A time dimension might contain a complete record of events within the business, broken down by hour, day, week, month, quarter, and year. 

Product Dimension

Product dimension might include information like a product’s name, description, color, and weight. 

People Dimension

The people dimension usually includes a customer’s name, age, location, and purchase history.

Why Dimension Tables Are Important

Dimensions provide structure for specific events and projects. Organizations often distribute dimension tables as templates for use across multiple reports. 

Without dimensions, companies would be unable to quickly query data and run custom reports on customers and projects. This is why they’re so important.

data dimensions and measures

Event Data Cloud: Powering Event Data at Scale

At the end of the day, an event is ultimately worthless if it can’t be acted on in a timely manner. Yet event data is very difficult to process at scale. Not surprisingly, this is why many companies are struggling in their data warehousing efforts.

Enter Scalyr’s Event Data Cloud, a purpose-built platform that businesses can use to ingest and query large volumes of data at scale in real time.

Event Data Cloud can help discover event and machine data and lay the groundwork for interactive analytics and data visualization. Taken together, this makes it the ideal tool for augmenting and enhancing data warehousing solutions.

Event Data Cloud can act as the main analytics engine for an application or service, enabling faster processing for repetitive and ad hoc analysis. In fact, many organizations are now using Event Data Cloud as a replacement for Elasticsearch to get the most out of all of their data.

Organizations can use this service to enhance security information and event management systems (SIEM) and security, orchestration, and response (SOAR); as a sink and persistent store for Kafka; and as a data pipeline to artificial intelligence and machine learning algorithms, among other use cases.

For further reading on this product, be sure to check out Scalyr’s comprehensive Event Data Cloud white paper, which goes into great detail about what event clouds are—and aren’t—and why companies should consider deploying this technology. 

The primer covers a variety of topics, including first-generation event data cloud solutions, examples of modern event data cloud solutions, overall technical requirements for an event data cloud, and much more. 

Take Scalyr’s Event Data Cloud for a Spin Today

Scalyr’s Event Data Cloud can help your business boost query speeds and become more event-driven. It can unleash the power of event data across a wide range of applications and services.

To see the full power of Scalyr’s Event Data Cloud in action, demo the service today.

This post was written by Justin Reynolds. Justin is a freelance writer who enjoys telling stories about how technology, science, and creativity can help workers be more productive. In his spare time, he likes seeing or playing live music, hiking, and traveling.