Your Vote Makes a Difference - Excelerator BI

Your Vote Makes a Difference

Microsoft has always been great at building and delivering world class robust commercial software, but it hasn’t always been great at listening to specific customer concerns about gaps in the software. All this has changed now, and Microsoft now has a clearly defined process to ask its customers what is important to them, and then prioritising these things for the new releases. Power BI (and Excel) both have specific user voice websites where anyone can submit their ideas for improvement.  In the case of Power BI (at least), the team DOES read the feedback and acts on the best ideas.

It works like this –

  • Any Power BI user can post an idea in the Power BI ideas forum –
  • The other users can then vote for the idea if they also find it useful.
  • The ideas that get large number of votes are considered by Microsoft for implementation.
  • Even the ideas that don’t get a large number of votes still get consideration.  More on that later.

Many ideas have been posted in the last 3 years and many of them have been implemented. Whenever I find that adding or enhancing a feature would useful, first go to and search for an idea that someone else has already created.  If I can’t find a similar idea already, I post a new idea. I also generally look at the other ideas regularly and vote for them if I also find them useful.

Below are the ideas that I have created that I think should be implemented. Yes, you got it! Your vote can make a difference to these changes being implemented.  Please take a look at these ideas and importantly please VOTE for them if you agree in their merit.

Conditional formatted measures using SWITCH

A current limitation in Power BI SWITCH measures is that they must have a single format. It is currently not possible to conditionally format the measure result based on any criteria – it is one single format only.
There are valid use cases where you may want to change the format of the SWITCH measure depending on the result.

Power BI DAX Book

IntelliSense Should Use Best Practice Column and Measure Naming Conventions

There is a clearly defined and accepted best practice for referencing columns and measures in the DAX language.
1. Column names should ALWAYS reference the table name.
TableName[Column Name]
2. Measures should NEVER reference the table name.
[Measure Name]
This is a fundamental concept accepted by everyone with knowledge of the DAX language (everyone I know anyway).
IntelliSense in the DAX sometimes follows this convention and sometimes it does not. I have created a short video (40 seconds) showing 3 examples.  It seems to fail when ever the column name and the table name have the same name, such as MyName[MyName of Col].
There are other places where column names are appended to measure names too, such as inside Quick Measures, and I have also seen Column Names used without prepending the table name.
This is an issue, and fixing IntelliSense so it always follows the best practice would enhance user experience.

Power BI Online Training

Improve Relationship Drag and Drop Experience

In Power Pivot for Excel, dragging and dropping columns to create relationships works like this.
1. Click and drag source column.
2. Hover over destination table.
3. The receiving column on the destination table is highlighted to indicate which column will receive the join.
In Power BI Desktop, item 3 above is missing. It makes it ambiguous how the join will be made (and a poor user experience). Adding a visual confirmation to the user which column will receive the join during drag and drop would be helpful.
Also, if you click or hover on the relationship, make the tables scroll so the user can see where the join is.

Read Only Access to Relationship View in Thin Workbooks

My definition of a thin workbook is when you connect Power BI to a tabular server. This can be SSAS Tabular or the Power BI Service (golden dataset as covered here
When you use Power BI Desktop as a thin workbook, you cannot see the relationships view or the data view. While this makes sense (they are not editable), it means that it is very difficult for someone to write local measures. It is not possible to accurately write a local measure unless the relationships between the tables are visible.
Please add the relationships view so the end user can see the table structure and assist in writing measures.

Link a Bookmark to Image with Spotlight Enabled

I would like to create a bookmark with a visual in spotlight mode AND provide a button to return to regular mode. Currently (as far as I can determine) I can place a visual in Spotlight mode and then add a bookmark for this. I can then link this “spotlight bookmark” to an image for the user to click. But then when Spotlight mode is on, there is no ability to click on an image/button linked to another bookmark that returns me to another view. I want to be able to click on a button (or image) to display a bookmark with a Spotlight highlight, and then somehow get back to the previous view or a specific bookmark.
This is currently not possible unless you do it directly in the bookmark pane. I would like to click on an image on the screen linked to Bookmark (or some other solution that is intuitive).

Mapping Visuals Should Send All Location Data to the Bing Service

If I add a list of State codes to the location card on a map, the results will normally not be accurate. The States selected could be from anywhere in the world – if you don’t add the country details first, you get inconsistent results. If I have the following Columns in my data model: Address, State, Post Code, Country, and assuming they are all categorised correctly using Data Category, then these columns should be sent to the Bing Service regardless if they are added to the visual or not. So if I put Address onto the map, I would like the full set of categorised columns that exist in my table to be sent to the Bing Service even though State, Post Code and Country are not actually placed in the visual.
This would greatly increase the usability and accuracy of the mapping visuals.

DAX Query Language Visual

There is a big value in creating a visual that can display the results of a DAX query.
This would give report writers very granular control to display records without having to use a table visual and a measure that triggers the display of the correct records.

When Cross Filter is Off, Chart Should Not Be Clickable

Currently when you turn off Edit Interactions for an object such as a column or bar chart, that chart will stop filtering any other items on the canvas. However when you click on one of the bars or columns, you can still select the item though it doesn’t do anything. This is confusing to the report users.

Use Power BI Desktop as a Local Server on a PC

There is an “Analyze in Excel” feature in the Power BI Service and I would like to see the same for Power BI Desktop.
There is already a “Connect to Data” button that connects to the service. You could enhance this to include a button that connects to a local instance of Power BI Desktop running on the PC.

Extend “Include In Report Refresh” Further

It would be great to allow the user to set parameters and events to control when a table is refreshed.
This will allow granular management of refreshing when a user needs to partially refresh a workbook.

The Small Stuff List (The SS List)

If there is a way to request small stuff that needs to be fixed, it can be reviewed by developers and if they agree, they can just fix the small stuff when working in the bigger stuff (assuming it is easy) or they can recommend it be promoted to an idea if it really is big stuff.
If the ideas are categorised into areas, the developers can easily find their own stuff, and also a way to remove things from the list once they are done.

Recently when I was in Seattle at the MVP Summit, I got chatting to one of the many great Microsoft Developers (Kay Unkroth).  Kay told me that people DO look at ALL of the ideas regardless of how many votes it receives.  What happens, when Microsoft is about to do some work in a particular area of the software, is that the team goes to and finds all the ideas that are relevant to that part of the software.  It makes sense to bundle up all the small stuff at the same time as you are about to work on a section of the code base.  This is a long way around saying that you should create ideas for everything you want to see developed.

If you find any or all of these ideas useful, please vote for them … Your Vote Makes a Difference!  And how about just browsing the top ideas and vote up the other ideas you think look worthwhile.

12 thoughts on “Your Vote Makes a Difference”

  1. Accepting user suggestions / input is a good thing, but …

    Why are you not using the UserVoice forum. That is where MS is directing all of it’s other user input.

  2. It would be interesting to see a report on how many ideas have been picked up and implemented. Seems it would be easy to do, and great marketing

    1. Funny you say that Dave. I am 75% through writing a follow up article where I will share how to keep track of the changes in votes over time. The same technique could be used to report on the status from idea to completed. Keep an eye out, probably next week.

      1. Will keep my eyes open for sure.

        Thanks again for a excellent session on Sunday at the Summit, I learned a lot.

  3. @Matt – Listening to Customers is great – But there has to be a higher benchmark
    If a request has to come from a customer for Providing a option to Change Font Color / Size of a Text Box to me does not indicate “world class or robust” . It means they released a half baked product/feature

    So when Avi Singh requests for allowing Excel to Connect to Power BI Data Model (Both Cloud and Local) and Microsoft conveniently listens to half the request – Some one like Matt has to create a separate “Idea”
    Don’t you think MS should have come clean and said this is something we can do but won’t – Because we will loose revenue

    1. I don’t think it is a revenue issue – JMO. I don’t think it is realistic to spend 5 years build everything needed to meet everyone’s needs *before* making the product available. The product would be crap if that were the approach as there would not be a good enough usage and feedback look. At the start of new software, the list of things to do includes *everything*. In the first 18 months of its life, I didn’t think Power BI was good enough for serious use. That is not the case now. It is not perfect of course. I don’t think it is fair to compare Power BI (3 years old) with Excel (33 years old). There will definitely be a difference with maturity. As for the direct access to the local instance of the PBIX file, this maybe all about to change. Kasper de Jonge announced at the MS Biz Summit this week that Microsoft now officially supports connection to a local instance of Desktop using DAX Studio. I don’t see any difference with doing it with DAX studio and Excel. I will follow up with Kasper to see if I can get this on the radar.

      1. @Matt – I am not saying they should wait for 5 years.
        But MS has so much of In-house experience that the Benchmark should be higher.
        Take my example of a the Text Box – Is this is new concept – NO – Are there a Dozen MS Products that have a Text box – YES (Excel, Word, PPT etc)
        So how can you ship a Text Box inferior to the Text Boxes in other Products that you have created

        Take another example of +/- buttons for Matrix – It has been in Excel Pivots since ages. So Excel Pivot has to be the Base benchmark for a Matrix in Power BI – You cant release an Inferior product / feature and then take customer suggestions and say we implemented them so pat on the back

        Same thing for almost every visual that was shipped in the initial releases of Power BI – They were Sub standard compared to their Excel counterparts for a Long long time – It is only in the last 6 to 8 months that things have started to get better

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top