Alteryx Insights

Analyze actual-versus-goal variance for a measure for a single period or non-time-series dataset, with optional breakdown. Use this to explain where actual performance is above or below goal. If breakdown is omitted, the tool automatically searches relevant segments and explains the strongest variance drivers.

insights__calculate_goal_variance

Analyze how a measure changed between two periods, with optional breakdown by segment. Use this to explain whether a measure increased, decreased, or stayed stable between two periods. If breakdown is omitted, the tool automatically searches relevant segments and explains the strongest drivers of change.

insights__calculate_change

Compare two queries and return correlation analysis for measures over the same dataset and time period. Returns a Pearson correlation between two measures over the same dataset and period.

insights__calculate_correlation

Calculate grouped SUM or AVG aggregations across all dataset rows matching the given filters. This tool does not provide a source link. Use for simple grouped summaries and exact top/bottom N by grouped total or average. Before calling this tool, first plan the complete data needed to answer the request, including all required aggregations, filters, and sort criteria. Consolidate those requirements into one request per dataset whenever possible; do not make separate calls to explore, sample, or retrieve individual aggregations. Make another call only when the planned request cannot provide sufficient evidence, such as when aggregation columns are skipped or truncated, or a genuinely different filter is required. For top/bottom N, request aggregations with a sort on the metric, then use the sorted aggregation groups. SUM and AVG results are returned separately so each can be sorted independently. For raw row-level lookups without aggregation, use get_dataset_rows instead. Use analytics tools instead for drivers, significance, trends, variance, outliers, correlation, or "why" questions.

insights__calculate_dataset_aggregations

Identify unusual changes in a measure between two periods, with optional breakdown. Use this to detect anomalies in change between periods. If breakdown is omitted, the tool automatically searches relevant segments to find the strongest outliers.

insights__show_outliers

Analyze the distribution of a measure for a single period or non-time-series dataset, with optional breakdown. Use this to explain total composition, including statistically significant high/low segment drivers. If breakdown is omitted, the tool automatically searches relevant segments and explains the strongest composition drivers. Not for exact ranked lists, leaderboards, or exact top/bottom amounts.

insights__calculate_totals

Identify unusual values in a measure for a single period, with optional breakdown. Use this to detect anomalies in totals for a selected period. If breakdown is omitted, the tool automatically searches relevant segments to find the strongest outliers.

insights__show_total_outliers

Analyze measure trends over time, including highs, lows, and forecast output.

insights__calculate_trend

Fetch filtered and sorted dataset rows. This tool does not provide a source link. Use for record-level questions, lookups, and row lists. Before calling this tool, first plan the complete data needed to answer the request, including all required columns, filters, and sort criteria. Consolidate those requirements into one request per dataset whenever possible; do not make separate calls to explore, sample, or retrieve individual columns. Make another call only when the planned request cannot provide sufficient evidence, such as when the row limit is reached or a genuinely different filter is required. Use analytics tools instead for drivers, significance, trends, variance, outliers, correlation, or "why" questions.

insights__get_dataset_rows

List datasets available in Alteryx One for analysis, one page at a time. Each dataset reports its current status, which tells you whether it can be analysed immediately or must first be prepared with prepare_dataset_for_insights.

insights__list_datasets