The Clearly Podcast

Dataflows in Power BI

Summary

This week we go deep into Dataflows in Power BI.

Dataflows are a method of building reusable content in the Power BI service that other users can use to help build other models. By using Power Query in the cloud, you’re able to perform transformations on one or many entities, storing those transformations as a cloud-based service. When you come to build or extend a model, the transformed entity is available, so you don’t need to recreate all of the steps to make the entity easier to use.

We talk through the benefits of using dataflows and who we think they’re for. We also discuss when you wouldn’t want to use this method, relying on the “traditional” desktop tools.

As we don’t see this as a replacement for Power BI desktop, at least anytime soon, we position dataflows as a tool that will benefit you in certain conditions. 

Documentation on dataflows can be found here https://docs.microsoft.com/en-us/power-bi/transform-model/dataflows/dataflows-introduction-self-service

If you already use Power BI, or are considering it, we strongly recommend you join your local Power BI user group here.

Transcript

Andy: Last week's recording was fantastic.

Tom: Without Shailan, you mean?

Andy: It’s great to have you back, Shailan. It wasn't the same without you, despite all the jokes with Tom.

Shailan: Yeah, I missed it. I was on a short holiday in Liverpool. Great city, nice weather, lots of shops, restaurants, bars, and the Beatles stuff. Highly recommend it.

Andy: Sounds lovely. I'll check it out when I get back to the UK.

Tom: Is this where we mention Liverpool City Council and the tourist board?

Andy: If only they'd sponsor us! Anyway, today we're talking about Dataflows in Power BI. Let's dive in.

Shailan: Dataflows in Power BI are like queries that fetch and transform data. You might be familiar with Power Query in Power BI, which lets you bring in and transform data.

Andy: When you say query, you mean Power Query, right?

Shailan: Yes, Power Query. It’s also available in Excel. But focusing on Power BI, once you've transformed data in Power Query, it loads into the Power BI model for visualizations.

Andy: What’s the advantage of using Dataflows?

Shailan: Dataflows are reusable and can be shared. You can create a dataflow to bring in data, transform it, and then make it available across the Power Platform, including Power Apps and other services. This reuse is a key benefit.

Tom: Exactly. Dataflows allow for reusability across multiple Power BI files, which means you don't need to repeat the ETL process for each report. However, it doesn't handle data modeling, which still requires Power BI Desktop.

Andy: So, it simplifies the process by allowing reusable data transformations. Customers can quickly adopt this to build models faster, especially with premium capacities.

Shailan: Yes, and it integrates well with the rest of the Power Platform, making data available for use in Power Apps and other services.

Andy: Are we moving towards using Power BI entirely in the cloud?

Tom: Dataflows are a step in that direction, but for now, data modeling still requires the desktop app. There are some transformations and AI components easier to handle in the cloud, but complex tasks might still need desktop functionalities.

Andy: Should organizations start using Dataflows now?

Shailan: It depends on their requirements. For reusable and shared data, Dataflows are excellent. For simple, one-off reports, Power Query within Power BI might suffice.

Tom: Early consideration of Dataflows in projects is beneficial, especially for centralized data management and reuse. But it's not yet the default for all scenarios.

Andy: Any final tips?

Tom: Think about reusability. If a data set needs heavy transformation and will be used in multiple reports, Dataflows are the way to go.

Shailan: Don't default to one approach. Evaluate each use case to decide whether Dataflows or Power Query fits better.

Andy: Great advice. Thanks, everyone. For more information, visit clearlycloudy.co.uk or clearlysolutions.net for the USA. Until next week, goodbye.

Tom: Cheers, Andy.

Shailan: Thanks, everyone. Bye.