Property "summarize by" vs. create a measure

[Nicole_PBI]

Hi all,

Assuming I want to display Revenue by a category, such as year. 

What is the difference between the following two options:

1. Set advanced property "summarize by" to Sum, and put the column on the visual.
   
2. Create a measure: my_measure = sum(Revenue), and put the measure on the visual.
   

Are there any differences from a result and design perspective? I'm aware that I'm more flexible by using a measure when it comes to row and filter context. But are there more reasons to use the measure instead of the summarized column? 

Thanks.

[Daniel_Reed]

Good distinction “Summarize by” works well for quick aggregation, while measures give far more control and consistency as models grow. When performance, reusability, or complex logic matters, defining a measure avoids hidden behavior and keeps results predictable, similar to how structured workflows are outlined Using measures also improves readability for others reviewing the model later.Overall, choosing the right approach early can save a lot of cleanup as reports scale.

[Anthony32]

This post raises a helpful distinction that affects how property data is interpreted in real situations. A brief example showing how summarized values differ from measurable outcomes could add clarity. Readers comparing assessment outcomes may benefit from resources like Putnam Tax Appeals when reviewing valuation results. Adding a short case reference would further support the explanation and practical use.

[Anthony_Wallace]

Good discussion—this topic comes up a lot when deciding how to control calculations versus raw fields. Using a measure often gives more flexibility, especially when context changes across visuals. I’ve seen similar clarity when reviewing real-world datasets like Hillsborough court documents where explicit logic helps avoid confusion. Clear definitions up front usually save time later when reports scale or get reused.

[HotChilli · Accepted solution]

From a result perspective : no difference.

From a design perspective : for simple, quick visuals , it's fine to use default summarisations .  For anything else, writing measures gives more control, can aid reuse (though be careful on this), and aids self documentation.

[Justin-Walker12]

This is a really helpful explanation! I’ve often wondered when it’s better to summarize data versus creating a custom measure, and your post clarifies it well. For anyone working with property data, tools can make analysis much easier and more accurate. I’ll definitely be experimenting with both approaches after reading this. Thanks for sharing these insights!

[HannahLee123]

This is a useful distinction, especially for anyone modeling reports with reusable logic. Relying on “summarize by” can work for quick views, but measures offer much better control and consistency as models grow. I’ve seen similar principles applied in data-heavy public records analysis, where clarity really matters, like how valuation data is explained through Escambia realty assessment resources. Defining measures upfront helps avoid confusion later and keeps reporting more scalable overall.

[Sophia_Baker]

This is a great point because choosing between summarizing a column and creating a custom measure really depends on the reporting goal. Summarization works well for quick visuals, but measures give far more control and accuracy as reports grow. I’ve seen similar clarity issues when working with datasets like martin property records where calculated logic matters a lot. Overall, understanding the difference early can save a lot of rework later and make dashboards more reliable.

[Nicole_PBI]

Thanks, that helps!