Create a Power BI report to analyze projects

In this task, we'll create a Power BI report based on the two lists. We'll bring the list data into Power BI Desktop and clean it up a little, do some basic data modeling, and create a set of visuals that tell us something about the data.

Quick review of Power BI Desktop

Before we dive into report creation, let's review Power BI Desktop. This is a powerful tool, with a lot of features, so we will focus on an overview of the areas that you will use in this task. There are three main work areas or views in Power BI Desktop: Report view, Data view, and Relationships view. Power BI Desktop also includes Query Editor, which opens in a separate window.

The following screen shows the three view icons along the left of Power BI Desktop: Report, Data, and Relationships, from top to bottom. The yellow bar along the left indicates the current view; in this case, Report view is displayed. Change views by selecting any of those three icons.

The Report view has five main areas:

1. The ribbon, which displays common tasks associated with reports and visualizations.
2. The Report view, or canvas, where visualizations are created and arranged.
3. The Pages tab area along the bottom, which lets you select or add a report page.
4. The Visualizations pane, where you change visualizations, customize colors or axes, apply filters, drag fields, and more.
5. The Fields pane, where query elements and filters can be dragged onto the Report view, or dragged to the Filters area of the Visualizations pane.

The Data view has three main areas:

1. The ribbon, which has the Modeling tab selected below. On this tab, you create calculated tables and columns, and make other changes to the data model.
2. The center pane, which shows data for the selected table.
3. The Fields pane, where you control how fields are displayed in your reports.

We don't use the Relationships view in this task, but you can check it out later after we bring the list data into Power BI Desktop.

In Query Editor, you build queries and transform data, then load that refined data model into Power BI Desktop. Query Editor has four main areas:

1. The ribbon, which has many options for shaping and transforming the data that you bring in.
2. The left pane, where queries are listed and available for selection, viewing, and shaping.
3. The center pane, where data from the selected query is displayed and available for shaping.
4. The Query Settings window, which lists the query's properties and data transform steps that have been applied.

Step 1: Get data into Power BI Desktop

In this step, we'll first connect to the two lists. Then we'll clean up the data by removing columns we don't need for our data analysis. We'll also change the data types on some of the remaining columns so that calculations work properly. For more information on getting and cleaning data in Power BI Desktop, see the Getting Data section in our Guided Learning course.

Connect to lists

1. In Power BI Desktop, on the Home tab, click or tap Get Data, then More…

   

2. In the Get Data dialog box, click or tap SharePoint Online List, then Connect.

   

3. Enter the URL for your SharePoint site, then click or tap OK.

   

4. If you get the following dialog box, make sure you're signed in with the right credentials, then click or tap Connect.

   

5. Select Project Details and Project Requests, then click or tap Edit.

   

   The lists are now displayed as tables in Query Editor.

   

Remove unnecessary columns from the tables

1. In the left navigation pane, select Project Details.

2. In the middle pane, select the FileSystemObjectType column, then click or tap Remove Columns.

   

3. Remove the two columns after the Id column: ServerRedirectedEmbedURL and ContentTypeId.

   Tip

   Use the Shift key to select both columns, then click or tap Remove Columns.

4. Remove all columns to the right of the PMAssigned column (a total of 22 columns). The table should match the following image:

   

5. Repeat the process you just went through, now for Project Requests: remove FileSystemObjectType, ServerRedirectedEmbedURL, ContentTypeId, and all columns to the right of the Approved column (a total of 22 columns). The table should match the following image:

   

Change the data type on Project Details columns

1. Select the ProjectedDays column, click or tap Data Type: Any, then Whole Number.

   

2. Repeat the previous step for the ActualDays column.

3. Select the ApprovedDate column, click or tap Data Type: Any, then Date.

   

4. Repeat the previous step for the ProjectedStartDate and ProjectedEndDate columns.

Change the data type on Project Requests columns

1. Select the EstimatedDays column, click or tap Data Type: Any, then Whole Number.

2. Select the RequestDate column, click or tap Data Type: Any, then Date.

Apply and save changes

1. On the Home tab, click Close and Apply to close Query Editor and go back to the main Power BI Desktop window.

   

2. Click or tap File, then Save, and save with the name project-analysis.pbix.

Step 2: Improve the data model

Now that we have the data from our lists pulled into Power BI Desktop, we'll move on to data modeling. Data modeling can be a time-consuming process, but we'll briefly show you some interesting things you can do to get more out of the list data in Power BI Desktop:

• Change how the two tables are related to each other
• Add a date table so we can make calculations based on weekdays
• Add calculated columns to calculate timespans between project milestones
• Add measures to calculate variance in projected versus actual days for a project

After these steps are complete, we can build visualizations that take advantage of the improvements to our model. For more information on modeling data in Power BI Desktop, see the Modeling section in our Guided Learning course.

Change table relationships

When Power BI Desktop brought the lists in, it created a relationship between them based on the Id column in both tables. The relationship should actually be between the Id column in the Project Requests table, and the RequestId column in the Project Details table. Let's fix that:

1. Click or tap the Data view icon.

   

2. On the Modeling tab, click or tap Manage Relationships. We'll stay on this tab in the Data view for all the data modeling steps.

   

3. Make sure the existing relationship is selected, click or tap Delete, then Delete again to confirm.

   

4. Click New to create a different relationship.

5. In the Create Relationship dialog box:

   1. For the first table, select Project Requests, and the Id column.

   2. For the second table, select Project Details, and the RequestId column.

   3. The screen should look like the following image. When you're ready, click or tap OK, then Close.

      

Add a date table to make date-based calculations easier

1. Click or tap New Table.

   

2. Enter this formula into the formula bar: Dates = CALENDARAUTO().

   

   This formula creates a table called Dates with a single date column. The table covers all dates from your other table, and it updates automatically if additional dates are added (i.e. if data is refreshed).

   This formula and the other ones in this section use Data Analysis Expressions (DAX), a formula language for Power BI and other technologies. For more information, see DAX basics in Power BI Desktop.

3. Press Enter to create the Dates table.

   

Add a calculated column to the Dates table

1. While still on the date table, click or tap New Column.

   

2. Enter this formula into the formula bar: IsWeekDay = SWITCH(WEEKDAY(Dates[Date]), 1,0,7,0,1).

   This formula determines whether a date in the Date column is a weekday. If the date is a weekday, the IsWeekDay column gets a value of 1; otherwise it gets a value of 0.

3. Press Enter to add the IsWeekDay column to the Dates table.

   

Add a calculated column to the Project Details table

1. In the right pane, click or tap the Project Details table, then New Column.

   

2. Enter this formula into the formula bar:

   ApprovedStartDiff = CALCULATE(SUM(Dates[IsWeekday]),
      DATESBETWEEN(Dates[Date],
         'Project Details'[ApprovedDate],
         'Project Details'[ProjectedStartDate]
     )
   )
   

   This formula calculates the difference in days between when a project was approved and when it is projected to start. It uses the IsWeekday column from the Dates table, so it counts only weekdays.

3. Press Enter to add the ApprovedStartDiff column to the Project Details table.

   

Add a calculated column to the Project Requests table

1. In the right pane, click or tap the Project Requests table, then New Column.

   

2. Enter this formula into the formula bar:

   RequestDateAge = CALCULATE(SUM(Dates[IsWeekday]),
      DATESBETWEEN(Dates[Date],
         'Project Requests'[RequestDate],
         NOW()
      )
   )
   

   This formula calculates the difference in days between when a project was requested and today's date (NOW()). Again, the formula counts only weekdays. This column is used to look for the project that has been pending the longest.

3. Press Enter to add the RequestDateAge column to the Project Requests table.

   

Add a measure to the Project Details table

1. In the right pane, click or tap the Project Details table, then New Measure.

   

2. Enter this formula into the formula bar:

   VarProjectedActual = DIVIDE(
       SUM('Project Details'[ActualDays]) - SUM('Project Details'[ProjectedDays]),
       SUM('Project Details'[ProjectedDays])
   )
   

   This formula calculates the variance between actual and projected days for a project. We add this as a measure, rather than a calculated column, so it returns the correct results regardless of how the data is filtered or aggregated in a report.

3. Press Enter to add the VarProjectedActual measure to the Project Details table.

   

Add a measure to the Project Requests table

1. In the right pane, click or tap the Project Requests table, then New Measure.

   

2. Enter this formula into the formula bar:

   MaxDaysPending = MAXX(
       FILTER('Project Requests', 'Project Requests'[Approved]="Pending"),
       'Project Requests'[RequestDateAge]
   )
   

   This formula finds the project that has been pending the longest, based on the calculated column we defined earlier.

3. Press Enter to add the MaxDaysPending measure to the Project Requests table.

   

Step 3: Create report visualizations

Now we're at the step that many people think of when they think of data analysis: creating visualizations so we can find patterns in our data. In this step, we'll create four visualizations:

• A column chart that shows projected days versus actual days on projects
• A column chart that shows the variance for each project
• A card that shows the project that has been pending the longest
• A table that shows the time between project approval and projected start date

After we've created these report visualizations in Power BI Desktop, we'll publish the data and reports to the Power BI service, so we can create and share dashboards. For more information on creating reports in Power BI Desktop, see the Visualizations section in our Guided Learning course.

Create a bar chart to show projected versus actual

1. Click or tap the Report view icon. We'll stay in this view for the rest of our time in Power BI Desktop.

   

2. In the Visualizations pane on the right, click or tap Clustered column chart.

   

3. Drag PMAssigned and Title from Project Details in the Fields pane to Axis in the Visualizations pane.

   

4. Drag ActualDays and ProjectedDays from Project Details in the Fields pane to Value in the Visualizations pane.

   

5. The visualization should now look like the following image.

   

6. Drag Status from Project Details in the Fields pane to the Filters area of the Visualizations pane, then select the Completed check box.

   

   The chart is now filtered to show only completed projects, which makes sense because we are comparing projected days to actual days.

7. Click the arrows in the upper left corner of the chart to move up and down the hierarchy of project managers and projects. In the following image, you see what the drill down into projects looks like.

   

Create a bar chart to show variance from projected

1. Click or tap on the canvas outside the visualization you just created.

2. In the Visualizations pane on the right, click or tap Clustered column chart.

   

3. Drag PMAssigned and Title from Project Details in the Fields pane to Axis in the Visualizations pane.

   

4. Drag VarProjectedActual from Project Details in the Fields pane to Value in the Visualizations pane.

   

5. Drag Status from Project Details in the Fields pane to the Filters area of the Visualizations pane, then select the Completed check box.

   

   The visualization should now look like the following image.

   

   You can see from this chart how much more variability there is for projects that were run by Irvin Sayers versus Joni Sherman. Drill in to see the variability by project, and whether the days projected was more or less than the actual days.

   

6. Before we create more visualizations, move and resize the ones you already created, so they fit side-by-side.

   

Create a card that shows the longest pending project

1. Click or tap on the canvas outside the visualization you just created.

2. In the Visualizations pane on the right, click or tap Card.

   

3. Drag MaxDaysPending from Project Requests in the Fields pane to Fields in the Visualizations pane.

   

4. Click or tap Format (paint roller), then set Border to On.

   

5. Set Title to On, then add the title "Max days pending approval".

   

   The visualization should now look like the following image.

   

   After we publish this report, we'll use this tile to trigger an alert if the maximum value for a pending project reaches a certain threshold.

Create a table that shows the time between project approval and projected start date

1. Click or tap on the canvas outside the visualization you just created.

2. In the Visualizations pane on the right, click or tap Table.

   

3. Drag PMAssigned, Title, and ApprovedStartDiff from Project Details in the Fields pane to Values in the Visualizations pane.

   

4. Drag ProjectedStartDate from Project Details in the Fields pane to the Filters area of the Visualizations pane, then select all dates except for (Blank).

   

5. Resize the columns of the table so you can see all the data, and sort by ApprovedStartDiff, descending. The visualization should now look like the following image.

   

6. In the Values area, click or tap the down arrow for ApprovedStartDiff, then click or tap Average. Now we can see the average duration between project approval and projected start date.

   

7. Click or tap the down arrow for ApprovedStartDiff again, click or tap Conditional formatting, then click or tap Background color scales.

   

8. Set colors for the Minimum and Maximum fields as shown below, then click or tap OK.

   

   The visualization should now look like the following image.

   

   As you can see, projects that Irvin Sayers runs tend to start a lot later after approval. There could be factors other than the assigned manager, but this would be worth looking into.

That brings us to the end of the report section, and you should now have a complete report based on data imported from SharePoint and cleaned up and modeled in Power BI Desktop. If everything went according to plan, your report should look like the following image.

Next steps

The next step in this tutorial series is to publish the Power BI project report and create a dashboard.

See also

• SharePoint integration scenarios