PowerBI.tips

Phil Seamark on Aggregation Tables

December 22, 2020 By Mike Carlo
Phil Seamark on Aggregation Tables

I was having a candid conversation with Phil Seamark from DAX.tips about Aggregation Tables. During that conversation, I was asking about patterns in using Aggregation tables. Within that 10 minute conversation I was blow away by all the possible patterns. Because of this, we pleaded for Phil to present these patterns to the Milwaukee Brew City User Group.

Patterns in Aggregation Tables

Watch the full webinar below from our user group:

Aggregation Patterns

While these patterns are described in detail here are the various patterns that can be used for Aggregation tables. Also, Phil includes a great introduction, found here. For each of these articles Phil describes proper usage for the pattern.

Horizontal Aggs

The Horizontal Aggs is typically developed using a time dimension. Thus, most aggregation tables would fit this pattern.

Example of a horizontal Aggregation table

Accordion Aggs

The Accordion Aggs are similar to a horizontal Agg. However, the time periods are not equal.

Accordion Aggregation table. Non uniform incremental aggregations.

Filtered Aggs

The Filtered Aggs would contain multiple Aggregation tables of the same data. But, each Agg table could contain different gains of data. For Example, data aggregated by Week, Month or Year.

Filtered Aggregation example.

Incremental Aggs

Finally, the Incremental Aggs. This type of aggregation would be used when aggregating transactions per day by store, and or product.

Incremental Aggregation where you can aggregated transactions over time

Thanks Phil

A special thanks to Phil for presenting. Since, we know you are a busy guy doing tons of great work. Thank you for taking time out of your day to present this wonderful topic. We hope you enjoy this exploration into Agg Tables.

Check out Phil’s other contributions

Phil has been quite active on the PowerBI.tips site. Check out all his other fun contributions.

Like and Follow

Previous

Power BI Direct Query Composite Models = Amazing

More Posts

Mar 13, 2026

What's Next? Building a 3-Year Power BI Roadmap - Ep. 510

Mike and Tommy map out what a practical three-year Power BI roadmap looks like as Fabric, AI, and enterprise data workflows keep colliding. If you're trying to decide what to standardize, what to experiment with, and what skills your team needs next, this episode gives you a grounded framework instead of hype.

Mar 11, 2026

Fabric Ideas That Stick, and What's Next - Ep. 509

Mike and Tommy tackle a mailbag question about submitting Fabric ideas, what makes a good idea stick, and where data engineering is headed next. Plus, Mike reveals he has 147 versions of Power BI Desktop downloaded.

Mar 6, 2026

Kicking Off Fabric the Right Way - Ep. 508

Mike and Tommy break down what it really means to start a Fabric implementation the right way, from setting expectations early to choosing practical first moves that won't box teams in later. They also cover a coming permissions change for Fabric data agents that makes semantic models easier for more people to use.