The Insights Platform and Insight Types in Tableau Pulse

When you create a metric in Tableau Pulse, you also automatically get the insights that Tableau Pulse detects for each metric.

The Insights platform in Tableau Pulse detects drivers, trends, contributors, and outliers for metrics. It proactively flags and describes insights that matter using natural language and visual explanations. The top insight for each metric is displayed with the metric.

Tableau Pulse also provides a path to further explore data by surfacing questions for the insights that it detects for a metric. This guided question-and-answer experience progressively reveals insights in the context of the metric. As you and others click through the suggested questions about the data, answers are revealed in easy-to-read charts with insights about the underlying data.

For more information, see Tableau Pulse: Proactive Answers to Your Common Business Questions with Automated Insights(Link opens in a new window).

Insight summaries highlight metrics of interest

When Tableau AI is turned on and you or others follow two or more metrics, Tableau Pulse provides an overview to help you quickly see the latest insights across your metrics of interest. This insights summary appears at the top of digests and in the Tableau Pulse home page.

Tableau Pulse looks across the metrics that you follow and leverages Tableau AI to summarize the most significant changes. Period Over Period Change and Unusual Change are the insight types considered for insight summaries.

Insight summaries use a large language model (LLM) to provide a personalized overview in plain language. Tableau AI is built on the Einstein Trust Layer, meaning it enables trusted, ethical, and open AI-powered experiences without compromising data security and privacy. For more information, see Tableau AI in Tableau Pulse(Link opens in a new window) and Einstein Generative AI for Tableau(Link opens in a new window).

Types of insights detected by Tableau Pulse

The questions users typically ask about metrics can be grouped into well-known patterns of analysis: descriptive, diagnostic, predictive, and prescriptive.

  • Descriptive questions: What happened to my Metric?

  • Diagnostic questions: Why did it happen?

  • Predictive questions: What is likely to happen next?

  • Prescriptive questions: What actions should I take?

The Insights platform in Tableau Pulse focuses on helping users answer basic descriptive questions they have about their metrics, such as:

  • How much has a metric value changed since the last period or the same period last year?

  • How is a metric trending over time?

  • Which members of a dimension contribute most to a metric value?

  • What other dimensions are driving a metric value in a favorable direction?

Insights in Tableau Pulse can alert users to hidden changes or anomalies in Tableau Pulse Metrics so they can better diagnose issues. For example:

  • Is the metric value higher or lower than normal?

  • Is the metric value unusually concentrated in a few entities of a dimension?

  • Has the trend of the metric changed recently?

  • Are there any unusually large records or outliers that are impacting the metric value?

Insight types in Tableau Pulse

The following insight types are used in Tableau Pulse:

Insight Type Description Configurable?

Record-level Outliers

Shows extremely high or low values for a metric, in the context of row-level values of the metric across a period of time.

Can be turned on or off in the Insights tab in metric definition settings. Requires selection of a record identifier field.

Period Over Period Change

Shows how a metric has changed between two periods.

This insight is displayed as part of every metric.

Always on by default

Tableau Pulse considers the Period over Period Change insight for insight summaries.

Top Contributors

Shows the highest values in a dimension for a metric within a given time range.

A top contributor is a dimension member that ranks in the top N in contribution to the scoped metric’s value, aggregated on a specified time range.

Always on by default

Tableau Pulse uses the Top Contributors insight in metrics for breakdowns.

Bottom Contributors

Shows the lowest values in a dimension for a metric within a given time range.

A bottom contributor is a dimension member that ranks in the bottom N in contribution to the scoped metric’s value, aggregated on a specified time range.

Can be turned on or off in the Insights tab in metric definition settings

Concentrated Contribution Alert (Risky Monopoly)

Shows when a small number of dimension members make up a majority (50% or more) of the contribution to a metric.

Can be turned on or off in the Insights tab in metric definition settings

Top Drivers

Shows values for dimension members that changed the most in the same direction as the observed change in the metric.

Can be turned on or off in the Insights tab in metric definition settings

Top Detractors

Shows values for dimension members that changed the most in the opposite direction to the observed change in the metric.

Can be turned on or off in the Insights tab in metric definition settings

Unusual Change

Shows when the value of a metric for a given time range is higher or lower than the expected range based on historic observations of the metric.

Always on by default

Tableau Pulse considers the Unusual Change insight for insight summaries.

Current Trend

Shows current trends to communicate the rate of change, direction, and fluctuations for the metric value.

Can be turned on or off in the Insights tab in metric definition settings

Trend Change Alert

Shows new trends that vary significantly from the current trend. This insight communicates the rate of change, direction, and fluctuations for the metric value.

Can be turned on or off in the Insights tab in metric definition settings

How Tableau Pulse generates and maintains trusted insights

Here are a few ways the Insights platform generates automated business insights users can trust:

  • Tableau Pulse Insights Service starts by using standardized, deterministic statistical models to detect facts about metrics that are guaranteed to be accurate. These facts act as the ground truth when generating insights.

  • Every insight that is generated is restricted to the data security context (such as RLS settings) of the user who made the request. This approach ensures users can only see the data they're authorized to see.

  • Analysts can enable or disable different insights being detected for a metric so they can control what is delivered to their users.

Here’s how Tableau Pulse brings it all together: the Insight platform statistical service uses the analytical context of the metric being followed or viewed to run automatic statistical analysis that generates facts about the metric. These facts answer the different questions using the user's data security context.

Insight summaries are generated using natural language grounded in statistical truths. The most relevant facts are processed by Tableau AI. These facts are used as ground truths to contextualize language generation. The resulting facts generated are bundled together and surfaced in several features throughout Tableau Pulse as insight summaries when Tableau AI is turned on for a site.

How the Insights platform determines relevance

To reduce noise, Tableau Pulse only surfaces the most relevant, useful insights—and avoids displaying noisy or spurious findings. The Insight platform considers the following factors to ensure the insights users see are relevant and useful:

  • The analytic context for insights is based on the Tableau Pulse metric definition. Unlike other solutions that look for insights across all columns in the data, the Insights platform restricts its analysis to the measures and dimensions referenced by the metric definition, as curated by analysts. In addition, only the filtered context of the metric user is viewing or following is considered when generating insights.

  • Insights are ranked based on impact to the metric. Each fact detected by the Insights platform is scored based on the impact it has on the metric value. Only the facts determined to be most statistically impactful to the metric value are returned first.

  • Feedback further personalizes insights. Users can provide thumbs-up or thumbs-down feedback on the insights they see to indicate whether the insights are useful. The Insights platform learns from this to further personalize the types of insights it shows to a user.

    Note: When the Personalized Insight Ranking site setting is turned on for your Tableau site, users can provide thumbs-up or thumbs-down feedback on insights. This setting is on by default. For more information, see Turn off Personalized Insight Ranking.

When applied in combination, these factors ensure that of all insights detected for a metric, only those found to be most useful are surfaced to users in Tableau Pulse.

Thanks for your feedback!Your feedback has been successfully submitted. Thank you!