Power BI key influencers: find what drives a metric up or down
Power BI key influencers is a visual that ranks the factors behind a metric. It analyzes the data, finds the factors that matter and lists them in order. For example, it can show which customer traits go with a low rating. This guide explains how to set the visual up, how to read both of its tabs and which limits Microsoft documents.
- Analyze and Explain by
- Influencers and segments
- Data needs and limits
Last updated
What does the Power BI key influencers visual do?
Microsoft describes two uses. First, the visual shows which factors affect the metric you analyze. Second, it contrasts how much each factor matters. So a team sees whether contract length or commute time has more to do with staff turnover.
The visual has two tabs. First, the Key influencers tab ranks single factors. Then the Top segments tab shows combinations of values, such as long-term customers in one region. A dropdown at the top sets the question, for example what influences a rating to be low.
It also reacts to the rest of the page. When a reader selects a slicer, a filter or another visual, the analysis reruns on that part of the data. Therefore one visual can answer the same question for each customer group.
Set up Power BI key influencers in four steps
These steps follow Microsoft's tutorial, which works in both Power BI Desktop and the Power BI service.
Add the visual
First select the Key influencers icon in the Visualizations pane. An empty visual then appears on the canvas.
Choose the metric
Next move the field you want to explain into the Analyze well, such as a customer rating.
Add possible drivers
Then move the fields that might matter into Explain by. You can add as many as you want.
Pick the outcome
Finally use the dropdown to choose the value of interest, for example Low, and read the ranked list.
How do you read an influencer?
Each bubble in the left pane is one influencer. For a categorical metric, the number shows how many times more likely the outcome becomes. Microsoft's example finds that customers in the consumer role are more than twice as likely to leave a low score.
Select a bubble, and the right pane shows the detail. A column chart compares every value of that field. The selected value appears in one color, and a dotted line marks the average of all the other values.
Numeric factors read differently. For a field such as tenure, the visual states how the likelihood changes when the field rises by a standard amount. A scatter chart with a trend line then shows the slope. If the link is not linear enough, the visual groups the numbers into bins instead.
Three settings that change the analysis
Microsoft documents these options, and each one helps the visual fit a different kind of question.
Analysis type
For a numeric field, switch between categorical and continuous analysis in the Format pane. Continuous asks what makes the number increase or decrease.
Expand by
This well sets the level of detail for a measure or a summarized column. As a result, an ID field adds rows without becoming an influencer.
Counts
Turn counts on in the Analysis card. A ring around each bubble then shows roughly how much of the data that influencer covers.
Which data do Power BI key influencers need?
The analysis runs at the level of the table that holds the metric. In Microsoft's example that is the customer table. Therefore every explanatory field must describe a customer. A field from a more detailed table causes an error until you aggregate or pivot it.
Volume matters as well. Microsoft advises at least 100 observations for the selected state and at least 10 for the states used for comparison. With only a few rows, patterns are hard to find, and the visual asks for more data.
Also check the table for a unique identifier. Without one, measures cannot be evaluated at the right level, so the analysis ignores them. So add an index column in Power Query when no ID exists, as Microsoft suggests.
Six documented limits to check before you rely on it
These points appear under considerations in Microsoft's article, so confirm them against your own setup first.
- DirectQuery models are not supported
- Live connections to Analysis Services are not supported
- Publish to web does not show the visual
- SharePoint Online embedding is not supported
- The analysis runs on a sample of 10,000 data points
- A categorical metric fails when implicit measures are discouraged
Power BI key influencers: questions from analysts
Why do Power BI key influencers show no results?
Microsoft names three causes. The metric may sit in both Analyze and Explain by, so remove it from Explain by. The explanatory fields may have too many categories with few observations. Or the data simply holds no meaningful pattern. For a numeric field, also try the continuous analysis type.
How does Power BI calculate a key influencer?
According to Microsoft, the visual runs a logistic regression for categorical analysis and a linear regression for numeric analysis. A statistical test then decides whether a factor counts as an influencer, with a threshold of 0.05. The number of data points matters too, so small groups often drop out.
Why does an obvious factor not appear?
Usually because the group is too small. The visual weighs the number of data points when it looks for patterns. So a role with a high share of low ratings can still be left out when few people hold it. The result could be chance, and the model treats it that way.
Why do influencers change when I add more fields?
Because the visual evaluates all explanatory factors together. A factor can matter alone but fade next to another one. For example, bedrooms may seem to raise a house price. However, once house size joins the analysis, bedrooms explain far less while the size stays constant.
Can Power BI Agency help us prepare data for driver analysis?
Yes. One of our consultants checks that your tables sit at the right level and carry a unique identifier, then builds the visual with you. We work remotely in your own Microsoft tenant and agree the scope in a written quote. Power BI Agency is a division of Sataware Technologies and independent of Microsoft.
Let's build your first dashboard together.
Tell us what you report on today. A consultant replies within 2 hours with next steps.
Trusted partners
Microsoft Solutions Partner · AWS Partner
Cloud & AI Platforms
Azure SQL, Microsoft Fabric and the data pipelines behind your reports.
AI Business Solutions
Power BI reporting on your Dynamics 365, Business Central and Power Platform data.
Security
Row-level security and Entra ID sign-in on every report and workspace.
Amazon Web Services
Power BI on your AWS data, with connectors for Redshift, Athena and S3.