November Full of Updates!

November is a mix of updates delivered with standard monthly schedule and those that were announced specifically during Microsoft Ignite. Let’s look collectively at all the features I consider interesting.

Is there anything for Power BI?

Maybe that’s a wrong question to ask. There always is, but this there anything that could make us happier. I think so.

Metric Sets

Feature that is of course in preview so far, but looks very good already, and may change the way we share the data today, especially for higher level management.  It provides a new of managing and consuming metrics – very good and very underrated feature of Power BI Service. You will be able to browse the metrics available in Metrics Hub. On top of a single value displayed you can find also a Metric Description, information when it was updates for the last time, from which Semantic Model it’s sourced, and view additional charts that are relevant for that Metric. It provides a great, clean, high-level overview of company’s KPIs and in many cases they could even replace (already outdated) Power BI Dashboards. Metric Sets will be a great addition not only for End Users, but also available to be re-used by content developers, ensuring data consistency across enterprise reporting.

Figure 1. Metric Sets (Source: Microsoft Fabric Blog).

If you would like to read more about this feature, here is a link with announcement: Announcing Metrics Sets are widely available in Preview | Microsoft Power BI Blog

Small multiples for New Card Visual

New Card Visual just like any other touched recently by Core Visual Team is great. They allow to finally make our Power BI Reports look modern. Still, there are many people who more than Visual side appreciate functionality of the chart. Previously, having multiple cards on a single page was avoided, not because they didn’t look good, but they were all separate object, requiring fair amount of time to render the visuals and calculate DAX measure. This of course impacted the performance, and small multiples finally solve that problem.

This is of course great benefit, but let’s not forget about other important factor – amount of work required to setup multiple card visual vs one visual using small multiples. No comparison here. I was waiting for it for quite some and hope to make a great use of this feature.

Small multiples with new card visual
Figure 2. Small multiples with new card visual.

Make sure to read full article on LinkedIn, released by PBI Core Visual team: New addition to Card visual: Introducing Small multiples! | LinkedIn

New Text Slicer

I was hesitating if I should include this feature here or not. On one side I am happy because I like using native visuals. It’s hard enough to build a good-looking Power BI Report, so I am not a fan of making it even more difficult by mixing visuals from different providers, where sometimes it’s hard to align the design between them. While level of visual customization is really exceptional, on the functional side is not that great. Comparing to other slicer visual available in the App Store, it doesn’t allow to switch between dimensions within single slicer and works with a single phrase only. And I used the term “phrase” on purpose, because when you type in more than one word it will search for those exact words in the same configuration. You are not allowed to look for more than one keyword at the same time, which makes the list of use case scenarios for this slicer quite limited.

This is just first iteration though, and I hope they will add more features to it. However, there is one more thing that worries me a bit. How many slicers we already have there? Each of them is made a separate visual in the Build Pane, and soon we will have a situation where there is a long list of visuals available to select, but half of them will be the different variations of Bar Chart and Slicer. Still, trying to stay positive and let’s see what the direction for New Text Slicer is, to make it really useful.

Fabric Data Pipelines Enhancements

Yes, this is not the mistake. Why am I talking about Fabric Data Pipelines within Power BI related section? It is because this particular Fabric workload got two important updates that are great for Power BI!

  1. Table and partition refreshes added to semantic model refresh – while this was available with external tools or APIs, it is now finally available for everyone. Something happened to your Incremental Refresh and you must get your data up to date – refresh only impacted partitions. Have you made changes to a single table only? – refresh only this table instead of full model.
    Refresh specific table/partition using Semantic Model refresh in Fabric Data Pipeline.
    Figure 3. Refresh a single table/partition using Fabric Data Pipelines.
  2. Schedule monthly refresh for your Data Pipeline – yes, this finally happened. Monthly reporting is still a thing, and I am surprised that till this point there is no out of the box solution to schedule monthly refreshes of Power BI Semantic Models. Now, combining Data Pipelines with action to refresh Semantic Model finally allows to have a monthly refresh without writing a single line of code. Is it then another great reason to start using Fabric? 🙂
    Schedule monthly refresh in Fabric Data Pipelines
    Figure 4. Schedule monthly refresh in Fabric Data Pipelines.

Important updates to Copilot Pricing

Some time ago I wrote a series of articles on Copilot, trying to assess its impact on the performance. I recommend to at least speed read them if you didn’t have a chance to work with Copilot yet:
 
What I highlighted there is that Copilot’s impact on performance can be significant, and what is even worse, you risk blocking your entire reporting when you overconsume Capacity. Which is not that difficult to imagine, considering unpredictable nature of Copilot related workload. That’s why I posted and idea, to Isolate Copilot operations to dedicated Capacity. During Microsoft Ignite we heard two important announcements for Copilot:
  1. Copilot’s Capacity cost is cut down by 50%
  2. It will now be possible to isolate Copilot’s workload to a dedicated Capacity, where no reporting is running.
Not only Copilot will no longer be that expensive, but you don’t have to risk anything and still enable Copilot to your End Users. This means on the other hand, that we can’t full ourselves anymore that Copilot is a free tool. Companies must make a conscious decision and plan the budget accordingly in order to enable Copilot for End Users. On the bright side, many big companies have some sort of Generative AI program, that maybe will be willing to bear the cost of Copilot Capacity 🙂
 

Fabric Databases

We can easily say that this was the most hyped announcement for Microsoft Fabric during Microsoft Ignite Event. I’ve seen a video where one of the DAX experts (I am not using names on purpose) was very happy about it, stating that this will close some gap for Power BI. That this is somehow compensating for the failed (or rather unfinished) concept of Datamarts. Datamarts were sold as a fully managed Azure SQL Database, so of course Fabric SQL Database seems to be a proper response to the needs. However, it was also supposed to be analytical solution, and for this we kind of already have Lakehouse + SQL Endpoint or Warehouse in Fabric. While I love the idea of having different workloads available in Fabric, here I have some doubts. Microsoft states that they are fulfilling the promise of delivering a unified data platform. The problem I have with this workload is that I am not necessarily sure how many people think that his is their dream coming true? Normally I could say that it’s not a problem at all, why not having different options available here, but let’s be realistic. Like everything that is developed today, it requires dedicated team effort and money spent on development. Money and time that could be spent somewhere else to have less but better workloads in Fabric, instead of having many that are OK.
As it usually happens for Preview Features, the list of limitations is quite big, with one of the biggest being that only Serverless type of Database is available. Which means that it will not probably be used for Critical Applications, where every millisecond counts.
However, I see a great potential of Fabric Databases within area of Business Applications. If you are using Power Platform today, you know that Dataverse is quite expensive solution and maybe we will slowly see it being replaced (at least to some extent) with Fabric Databases that are much cheaper.
 

OneLake Catalog

Maybe not that spectacular update but very important from perspective of Data Governance. Whatever contributes to discoverability of the data and helps to manage it more efficiently is a great update to me. Catalog can be browsed by Domains, Workspaces and even Tags, even though they are currently in Preview mode. Items display many useful information like description, sensitivity labels, endorsement. You can browse the tables available in given artifact, see the Lineage View, when data was updated for the last time, monitor refreshes… and so on. You have available options for data exploration, create a report based on the data. Capabilities are limited so far for Lakehouses, too bad there is no option yet to start working with this data directly from Catalog, but I hope something will pop up soon:
 
A screenshot of a computer Description automatically generated
Figure 5. OneLake Catalog in Fabric.
 

Tenant Switcher

Great update if you are external consultant handling multiple clients, or you are a company that is sharing Power BI Reports outside of your organization to Guest Accounts. In such cases, handling the multitenant access was difficult. Now it’s very easy to switch between the content as Microsoft introduced the Tenant Switcher:
Tenant Switcher in Microsoft Fabric
Figure 6. Tenant Switcher in Microsoft Fabric.
 

Conclusion

There is a lot of interesting features coming to Power BI and Fabric. Despite some small shortcomings, I think it’s going in good direction, and I hope, that this trend continues. I wonder what your favorite November updates are, do you agree with my picks, or you would like to add something else?
 
As always, thank you for staying till the end and see you in next article 🙂
Picture of Pawel Wrona

Pawel Wrona

Lead author and founder of the blog | Works as a Power BI Architect in global company | Passionate about Power BI and Microsoft Tech

Did you enjoy this article? Share with others :)

5 4 votes
Article Rating
Comments notifications
Notify of
guest
0 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments

Related Posts