How to use Insights

The Insights feature provides automated, AI-powered data variance and contribution analysis. It identifies the Dimensions and Properties driving changes in your KPIs and Metrics, providing visualizations and written explanations to enhance decision-making.

How Insights can help you

Here are three use cases Insights can help with.

Use Case 1: Variance Analysis

Teams frequently need to understand why performance has changed over time or across different versions. Instead of manually filtering through reports and pivot tables, Insights automatically identifies the top contributors of variances—whether it’s a shift in revenue, costs, or headcount. This helps Members quickly spot trends and make data-driven decisions.

Variance analysis can be applied across different timeframes and versions. Examples include:

Period-over-Period Analysis: Identify trends by comparing Month-over-Month (MoM), Quarter-over-Quarter (QoQ), or Year-over-Year (YoY) performance.
Version Comparisons: Track deviations between Actuals vs. Budget, Actuals vs. Forecast, or Forecast vs. Forecast.

Typically, Members can find the following from Insights:

Revenue: A sudden decline in total revenue needs further analysis. Insights helps users break down revenue shifts by region, industry, or customer type to pinpoint whether specific countries or segments are underperforming.
Cost: A significant increase in operational expenses requires investigation. Insights identifies changes in fixed vs. variable costs and detects whether increases stem from new vendors, increased service costs, or unexpected spikes in specific expense categories.
Gross Margin / Profit: If gross margin is fluctuating, users need to understand the root cause. Insights breaks down profitability by product category, helping to identify areas where margins are compressing.
Headcount and Compensation: If salary costs are misaligned with budget expectations, Insights can highlight which departments are contributing to the variance—whether due to delayed hiring, structural salary adjustments, or workforce reductions.
Marketing and Advertising Spend: A sharp rise in marketing expenses warrants closer examination. Insights enables users to analyze which campaigns, regions, or channels are driving the variance.

Use Case 2: Drill-down analysis

When analyzing high-level Metrics in an aggregated report, Members often need to identify the key drivers behind a change. With Insights, instead of having to manually select a specific Dimension to drill into, Members receive a breakdown of the most relevant Dimensions and top contributing factors. Insights removes the need to guess which breakdown to explore. Examples include:

A finance team is reviewing total revenue at a regional level and needs to see which countries or segments contributed the most to an increase or decline.
An operations team notices an unexpected increase in total expenses and wants to uncover whether it is driven by specific vendors, cost centers, or other dimensions.
A product team is examining gross margin shifts and wants to identify which product categories or business lines had the biggest impact.

Use Case 3: Data Validation

When importing or transforming data, Insights checks for consistency, highlighting unexpected deviations. For example:

Load validation: When loading new data, unexpected deviations in key business metrics can go unnoticed, leading to incorrect decisions. Insights helps automatically detect anomalies, such as missing vendors, significant discrepancies between periods, duplicate transactions, or unexplained spikes.
Transformation validation: If data transformations are applied across many Metrics, verifying the accuracy of the calculations can be time-consuming. Insights can automate the process of detecting inconsistencies by comparing initial, intermediate, and final Metric values.

Before you begin

Metrics’ AI set-up

Workspace Administrators need to set up the following for Insights. If you have already set this up for the Data Assistant, you can skip to the next section:

AI activation. Ensure AI has been activated as set out here.
Enable Metrics for AI. See this article for selection tips and guidance.
Configure AI Metrics and Dimensions. Optimize the below using the guidance in the links:

Prepare your Boards for Insights

Members can get Insights from Boards’ KPIs. To set this up, choose KPIs that:

originate from AI-enabled Metrics

measure either variance or contribution

The below table gives information on these two analysis types:

Analysis type	Description	Type of Views Supported	Sample prompt
Variance	Explains factors driving variance (e.g. actuals vs. budget, current vs. previous year)	Two cells	“Compare revenue year 23 to year 24”
Contribution	Drills down into Items driving the KPI	One cell	“What was my revenue last quarter"

Customize Insights results

You can fine-tune Insights results by defining key business-critical Dimensions:

In your Application, select the three dots at the bottom of the sidebar and then AI Customization.
In Allowed Metrics, open the AI analysis dropdowns by the relevant Metrics (see image below).
Select your preferred Dimensions.

Grant the AI analysis permission

Members need the AI analysis permission to use Insights. See this article for more information.

Access rights

The Insights feature only provides responses that the Member has the right to see.

For example, if a Member with access to a single country selects Get Insights, it returns Insights just for that specific country.

You can define Members’ access rights from the Roles, permissions and access page. See this article for more information.