Change analysis

With SpotIQ change analysis, you can compare any two data points in a visualization for change and identify key change drivers from the underlying attribute columns.

SpotIQ change analysis compares data points of simple or complex measures using these aggregates:

  • Sum

  • Count

  • Sum over sum, which generates 'what-if' percentage insights

  • Average, which generates 'what-if' percentage insights

  • Other functions which use a 'versus' analysis to show the absolute change grouped by different drill attributes

Contextual change analysis

ThoughtSpot supports contextual change analysis for area, line, column, bar, line-column, pie, and KPI charts. You do not need SpotIQ privileges to run contextual change analysis.

To perform this analysis, follow these steps:

  1. On an area, line, column, bar, line-column, pie, or KPI chart, hold down the command key (MacOS) or control key (Linux or Windows).

  2. Select the two columns or data points you plan to compare.
    Note that if you are analyzing a KPI chart, you must either click the percent change label or select two points from the sparkline visualization. You cannot run change analysis on a KPI chart with no sparkline visualization.

  3. Right-click to see the dropdown menu, and select Run change analysis.

  4. The SpotIQ window modal appears with the results of your analysis. By default, SpotIQ picks the top 5 most relevant columns, based on what it learned from your past activity.

    Contextual change analysis window

    To select other columns to analyze, click Customize attributes in the lower left corner of the window and select the desired column from the window that appears.

    Each colored section of a SpotIQ insight represents a statistically significant outlier among the values of your chosen columns. For more information on how statistically significant outliers are calculated, see Z-scores.
  5. You can now examine each visualization of the analysis for details of comparison.

    In some instances, ThoughtSpot may point out that certain attribute values account for more than 100% of the total change. This is generally for cases where attribute values contribute to a trend that is opposite to the total change (positive change when the overall trend is negative, for example). You may see cases, for example, where two attribute values account for a 300% increase in sales, while the remaining attribute values indicated account for 200% of the total decrease.
  6. You can add the specific visualization to an existing Liveboard, or create a new Liveboard and add it there.

  7. You can run another round of SpotIQ analysis, download the data as a CSV file, edit this visualization, or make a copy of the visualization.

How SpotIQ calculates outliers


A Z-score represents the number of standard deviations from the mean a data point is. Z-scores are a way to compare results from a test to a “normal” population. The basic Z-score formula for a sample is: z = (x – μ) / σ where

  • σ is the standard deviation

  • μ is the mean

  • x is the data value.

Currently, any value outside μ +/- N * σ (i.e. Z-score less than or greater than N) is an anomaly. N is determined based on the number of rows. It is chosen such that the max or min has < 0.1 probability of crossing the threshold. SpotIQ uses the following table to approximately choose the value of N, that is, how many standard deviations we want to be able to use as threshold:

Cardinality N (number of standard deviations from mean)

< 20


< 500


< 10000


>= 10000



  • Change analysis is not yet supported on measures that are created using group_* formulas.

  • Change analysis is not supported on queries based on the "growth of" keyword.

  • Change analysis is not supported on queries based on the "versus" keyword.