Categorie op Actionable BI

Hidden Gem: Power BI mobile reports

I am not the biggest fan of Power BI dashboards. They do look amazing navigating via mobile phone and are a great way to combine insights into multiple reports from one view but in my opinion they lack certain functionality to be used as a default.

What would make Power BI dashboards great

There are some basic functionalities that would improve the usability for Power BI dashboards in a great way. I’ll briefly summarize my view on these changes.

  • Updatable pinned visuals
    You decide your new report is finished. You pin a visual to a new or existing dashboard and you add the visual to the mobile layout. In the end you notice a filter has been applied during the pin-action. This requires the visual to be pinned to the dashboard and mobile view again. And I don’t mention the resizing and renaming in the dashboard. Change in colours, labels, set slicers etc require pinned items to be re-pinned. This makes the Power BI dashboard way to solid.
  • Global slicers
    Looking at the Power BI dashboard it could be that several visuals have a different granularity in for example years that are shown. To create a single glass to interpretate the data on the same granularity it would be an amazing addition to have something like global slicers that work over reports. So a global slicer for “Year” would filter all visuals to the same period. A good starting point would be to have the Edit interaction option from Power BI desktop
  • Multi-pin in mobile view
    A mobile dashboard view can be styled in a great way by using titles and images. If you want to add images to this mobile view you need to have them in your dashboard and even if you want to use the same image, like a logo, multiple times you need to have that image in your dashboard mulitple times. An option to multi-pin the same visual would be a great addition to the mobile view but also keep the original dashboard cleaner
  • More flexilibity in dashboard layout
    The Power BI dashboard can be improved with having more flexilibity in the dashboard tile layout. Being able to change the width and height without restrictions, default settings for titles, subtitles, font etc and last but not least the ability to add content items that are not shown like images that should only be visible in the mobile view

Eventhough I see a lot of improvements I still fancy the easy and intuitive way a dashboard can be created in very little time. In my day to day job I see a lot of businesses using a default approach to Power BI content. This starts with Power BI desktop, Power BI service (report), Power BI service (dashboard) and finally sharing it with collegues. Probably this has lots to do with dashboard sharing being the only option in the beginnen for Power BI.

User navigation

In a situation where a dashboard is shared with an end-user this person can directly go to powerbi.com to navigate to the shared dashboard. After selecting a visual the user is directed to the underlying report for further analysis. This works fine and the user can get to the report fairly quickly.

If the user navigates to the dashboard via the Power BI mobile App this experience is slightly different. I really fancy this dashboard view because it has lots of space with enough room for clear visuals.

Fig 1

This dashboard view is read-only and after selecting one visual it will open to full screen. The user can see the data labels here or choose to annotate and share his thoughts (Fig 2).

Fig 2

The interactivity on this visual is limited to this visual only. If this users want’s to interact with other visuals it can navigate to the underlying report (Fig 3) by selecting the little barchart in the top right corner.

Fig 3

Eventhough phones are getting bigger and bigger still this report view is not very user friendly. Especially when there are lots of visuals and slicers on the canvas.

Power BI Phone layout

From within Power BI desktop each report page can have a mobile phone layout. This can be selected from the View tab on the ribbon.

Fig 4

By dragging & dropping items from the right panel the mobile view can be created. Yes it’s true that multi-pin should be a great addition here also!

After saving and publishing the report this mobile view can now be accessed by selecting the litle bar chart on the top right from Figure 2.

Fig 5

Now also the report is readable on mobile devices and the user is able to interact between visuals as shown in Figure 6. This improves user experience greatly and data analysis on their mobile phone.

Fig 6

Does this replace the Power BI dashboard?

The Power BI dashboard still has some great features like subscriptions, alerts and multi-report visuals but if you don’t use these this mobile view could replace the Power BI dashboard. You can also directly share a Power BI report without having to create a Power BI dashboard.

Conclusion

I see lots of organizations using the Power BI mobile app without configuring their reports for mobile consumption. It takes a little bit more time to create a mobile view for every report page but it enhances the user experience on mobile greatly!

Dynamic titles in Power BI

Since the April release of Power BI desktop it is possible to use dynamic titles in your visuals. This can be super handy when there is a need to explain more than just a static title. For example a static title would be “Sales over time” and a dynamic variant could be “Sales between jan 2019 and july 2019”.

To get to this option select a visual, navigate to the title menu and select “Conditional formatting” from the three dots dropdown.

In the popup that appears you can select an attribute. This can be either a column or a measure but it must be from a Text data type. If you select a column you must specify the summarization because this dynamic title can only handle a single value. To get started it is important to know what context is applied to the visual and how we can access attributes that will create our dynamic title.

Evaluation context

In DAX there is always a evaluation context. This is the context that affect a DAX expression (calculated column, measure) when it is executed. Most important is that you have an understanding about filter context, row context and context transition. You’ll start exploring these kinds of contexts whenever you build a measure that does not return the value you would expect. To explain these contexts I prefer referring to the best explanation that can be found here. In this blog I will refer mainly to the evaluation and filter context.

Getting started

First lets start with a simple dataset and some visuals.

Fig 1
Fig 2

We’ve got some data for several months and two products. The title is set to “Sales over time” and this should have an indication from the period that data is shown, lets say “Sales from 1-1-2019 until 1-6-2019”.

To do this we need to be able to grab both dates from the dataset. This can be done if we know what the evaluation context for this visual is.
The evaluation context is the environment in which a formula is evaluated and this context changes through slicers, filters, rows and columns in visuals. For this visual there is a filter context at each data point (every date) but if we look at the chart itself we could refer to the title as being the “total” value. There is no filter set by any slicer, filter, row or column value. If you look back at the data table (Fig 1) you can see that there are no rows excluded by matching the values in the visual.

Building the dynamic title

Create a new measure from the field list. To create the dynamic title the below DAX formula would do the trick.

DynamicTitle = 
VAR FirstDt =
    MIN ( Data[Date] )
VAR LastDt =
    MAX ( Data[Date] )
RETURN
    "Value from " & FirstDt & " until " & LastDt

Because dates are chronologically ordered we can grab the first date by selecting the minimal value and the max date by selecting the maximum value. I prefer to use variables because this enhances readability and has soem benefits in only calculating once.

The final line builds the dynamic title resuling in the desired outcome.

We could enhance the title in combination with the Product slicer. A slicer changes the filter context so whenever one product is selected we could use a MAX() MIN() FIRST() LAST() or any other summarization to select this product the same way as with the date but we could also concatenate the selected products by using SELECTEDVALUE()

Title = 
VAR FirstDt =
    MIN ( Data[Date] )
VAR LastDt =
    MAX ( Data[Date] )
VAR Product =
    SELECTEDVALUE (
        Data[Product];
        CONCATENATEX ( VALUES ( Data[Product] ); Data[Product]; "," )
    )
RETURN
"Value from " & FirstDt & " to " & LastDt & " for product (" & Product & ")"
Fig3 The new title with product A selected
Fig 4 With both products selected

Conclusion

Dynamic titles are a great way to give more information about what is displayed in a visual. Whenever you start with dynamic titles you should consider when this title becomes a paragraph that creates information overload. The basic idea should be that it helps understanding the visual.