What is a Streaming Dataset in Power BI?

Your reports are snapshots in time, telling you what happened yesterday or last week, but streaming datasets in Power BI change the game by showing you what's happening right now. They turn your dashboard from a static summary into a live, breathing view of your business operations. This article will break down exactly what streaming datasets are, the different types available in Power BI, and how you can start using them to monitor your business in real time.

Graphed

Still Building Reports Manually?

Watch how growth teams are getting answers in seconds — not days.

Watch the Demo

What is a Streaming Dataset in Power BI?

A standard Power BI dataset pulls data on a schedule - you set it to refresh every hour or once a day. A streaming dataset, however, doesn't wait for a schedule. Instead, data is continuously pushed into it from a source, allowing you to see new information within seconds of it being generated.

Think about the difference between reading a newspaper printed once a day versus watching a live news ticker on TV. The newspaper gives you a comprehensive but dated report. The news ticker streams information as events unfold. Standard Power BI reports are the newspaper, dashboards with streaming datasets are the live ticker.

With a streaming dataset, there's no need to manually refresh or wait for the next scheduled update. The data source - whether it's a sensor, a website, a sales system, or a social media feed - sends data directly to your Power BI dashboard tile the moment an event occurs. This creates a powerful, low-latency connection between your business activities and your analytics.

Graphed

Still Building Reports Manually?

Watch how growth teams are getting answers in seconds — not days.

Watch the Demo

Why Does Real-Time Data Matter?

For many business operations, speed is everything. Waiting an hour for a data refresh can mean missing a crucial opportunity or failing to catch a problem before it escalates. Here are a few practical scenarios where real-time data is invaluable:

• Monitoring a Sales Flash Sale: See which products are selling fastest minute-by-minute and adjust promotions on the fly.
• Tracking a Marketing Campaign Launch: Watch sign-ups, clicks, and conversions in real time to see if your new ad creative is hitting the mark immediately.
• Social Media Engagement: Monitor mentions and sentiment live during a product launch or event to respond to customers instantly.
• Website Performance: Keep an eye on server response times or error rates to catch technical issues before they impact a large number of users.

The Three Types of Real-Time Datasets

Power BI offers three different approaches to handling real-time data, each with its own pros and cons. Understanding the differences is crucial for choosing the right one for your specific needs.

1. Push Datasets

A Push dataset is the most common and flexible type. When data is sent to a Push dataset, Power BI stores it in an underlying database. This is a key distinction: the data is saved and accumulates over time.

• How it works: Your application pushes data to a Power BI API endpoint. The data is stored, and your dashboard visuals update to reflect the new information.
• Best for: Situations where you need both a real-time view and the ability to perform historical analysis. Since the data is stored, you can build standard Power BI reports on top of it, slice and dice by date, and explore trends over weeks or months.
• The Catch: The live dashboards don't update instantly. While the data is pushed in real time, the dashboard tiles are typically on a refresh cycle themselves (e.g., every 15 minutes by default, though data arrives faster). It’s "near real-time" but not sub-second instantaneous for visuals.

2. Streaming Datasets

A true Streaming dataset is all about speed. Unlike a Push dataset, data sent here isn’t stored permanently in a Power BI database. Instead, it is held in a temporary cache for a short period (around one hour) to populate visuals.

• How it works: Data is streamed to Power BI and is immediately visualized on a dashboard tile. The visuals update in seconds.
• Best for: Pure monitoring scenarios where you only care about the present moment. Think monitoring factory sensor temperature, server CPU usage, or website traffic volume right now.
• The Catch: Because the data is not stored historically, you can't create traditional Power BI reports from it. You can't filter by date ranges, build complex charts comparing past performance, or ask questions of the data beyond what you see at that instant. Their use is limited to live dashboard tiles.

Graphed

Still Building Reports Manually?

Watch how growth teams are getting answers in seconds — not days.

Watch the Demo

3. Hybrid Datasets (Using PubNub)

What if you want the best of both worlds? The blinding speed of a Streaming dataset and the historical storage of a Push dataset? That’s where a Hybrid dataset comes in. This is achieved by using a PubNub streaming source when you set up your model.

• How it works: You configure a single dataset that functions as both a Streaming and a Push dataset. When data comes in, Power BI uses the streaming component to deliver sub-second visual updates to your dashboard while simultaneously using the push component to store that same data in its database for historical analysis.
• Best for: The most demanding real-time scenarios where you need instant alerts and the ability to perform deep-dive analysis later on. For example, tracking financial transactions where you want to see them happen live but also need to run reports on transaction history.
• The Catch: Setting up a PubNub streaming source is more involved than the other two types and requires an understanding of the PubNub platform.

Quick Comparison Table

How to Set Up a Streaming Dataset in Power BI

Creating a streaming dataset is surprisingly straightforward. You don’t need to connect to a source, instead, you define the structure (the schema) of the data you expect to receive, and Power BI provides you with an API endpoint to send data to.

Let's walk through creating a simple Push dataset to monitor website signups in real time. We're choosing Push because we want to see live signups while also being able to analyze signup trends over time.

1. Create the Dataset: In your Power BI workspace, click + New in the top-left corner and select Streaming dataset.
2. Select Your Source: Choose API as your source and click Next. This is the most universal method.
3. Define Your Values:
4. Click Create: Power BI will now generate a unique Push URL and a JSON snippet showing the required data format. Your work in the Power BI interface is done!

That Push URL is the endpoint you'll use in another application (like a simple script, an automation tool like Zapier, or your website's backend) to send data. Whenever your app POSTs new data to that URL in the correct JSON format, it will appear in your dataset, ready to be visualized on a dashboard.

{
  "signup_email":"Test@example.com",
  "location":"USA",
  "signup_time":"2023-11-21T10:00:00.000Z"
}

Graphed

Still Building Reports Manually?

Watch how growth teams are getting answers in seconds — not days.

Watch the Demo

Bringing a Streaming Dataset to Life on a Dashboard

Once your dataset is created and data is being pushed to it, you can create a new dashboard and add tiles that visualize this data.

Go to your dataset, create a basic report with a card visual showing the count of signups and a map showing location. Then, pin these visuals to a new dashboard. Now, as new signups are pushed to your API endpoint, you'll see the numbers on your card increase and new points appear on your map, giving you that 'live ticker' feel for your business's growth.

Final Thoughts

Streaming datasets in Power BI are one of its most powerful features, allowing businesses to move beyond stale, historical reports and embrace real-time decision-making. By choosing the right type - Push, Streaming, or Hybrid - you can create live dashboards that monitor everything from marketing performance to operational health as it happens.

Setting this all up can still require some technical comfort, connecting different services and understanding APIs. That's why we built Graphed to be the simplest path to real-time analytics. We handle the complexities of connecting directly to all your data sources - like Google Analytics, Shopify, and Salesforce - and an AI data analyst to create constantly updating, live dashboards for you in seconds using just plain English. No JSON, no API endpoints - just answers.