Small Multiples - Wrong data labels

[agnesj]

Dear Community, 

I am having an issue with the data labels of a Stacked Column Chart with small multiples. 

I am showing side by side sales data for 2024 and 2030 using small multiples, with Region as a Legend and Brand as x-axis. So I basically want to show the sales volume for 2024 and 2030 for each brand, with a split by Region inside of each bar. 

I want to add the share of sales volume of each Region for each brand. 

I used the following calculated measure : 

Market Share Country by Indication =

VAR countryms= CALCULATE(sum('Table 1'[Value]))

var allcountries= CALCULATE(sum('Table 1'[Value]), REMOVEFILTERS('Table 1'[Global Region]))

RETURN DIVIDE(countryms,allcountries,0)

This works as a tooltip, as shown below, but not as a data label (Detail). The data labels are basically the share of the bottom country for each bar. For example, for the first bar from the right (Brand 3, year 2030), 2% is the share of the yellow region. But what should appear is 49% for the Red Region. 

This measure works fine when there is no small multiples. 

Is this a bug by Power BI, and does anyone know how to resolve it ? 

Thank you very much ! 

[danextian]

Seems to be  a bug.  Revenue Percent to Category Total should be the data label but it is picking up that of the very first legend value which in my case is APAC.

[agnesj]

Hi @danextian ,

Yes I'm getting the same results, it seems like a bug.

Would you have an idea of a workaround ? 

[v-sgandrathi]

Hi @agnesj,

Thank you for confirming. Based on feedback from multiple users, it appears this issue is a limitation or bug in how Power BI currently renders data labels within small multiples for stacked visuals.

Possible Workaround:

While awaiting an official resolution, here are some temporary solutions you may find helpful:

Leverage Custom Tooltips:
Since your DAX measure functions correctly in the tooltip, consider creating a detailed tooltip page to display the accurate market share breakdown per region. This allows users to view the correct data on hover without relying on the data labels.

Split the Visuals:
Alternatively, you can duplicate the stacked column visual and apply manual filters by year instead of using small multiples. Although this requires additional layout work, it ensures more accurate label rendering.

Use Card Visuals for Labels:
For greater flexibility, you may create separate card visuals to calculate and display market share by region and brand. While this approach is more manual, it allows full control over label placement and content.

For a long-term fix, we recommend raising a support ticket with Microsoft Power BI Support.
How to create a Fabric and Power BI Support ticket - Power BI | Microsoft Learn

Thank you.

[amitchandak]

@agnesj , I usually prefer separate dimensions in such cases. So that removefilter only work with that dimension. Move date/Year and region on dimension and create star schema and adjust the formula.

Or use the approach of new feature as explained by Guyinacube 

https://www.youtube.com/watch?v=gTZ8YKW425s

[agnesj]

Hi @amitchandak ,

Thanks for your answer, but I tried your suggestion and it doens't seem to work. It seems like it's a bug.