Note: The Insights tab takes up to 24 hours to generate insights, with insight cards refreshed once every 24 hours.

How does Survey Insight's work?

The working of Survey Insights can be explained with the help of the following steps:

1. Get Data: When you create an insights report, all survey questions and responses are automatically shared with Sprinklr AI+ to provide full context. You can’t select specific data by default, the entire dataset is sent to the AI.

2. Analyse Data: Sprinklr AI+ runs text analytics, regression, correlation and descriptive analytics on your data to identify patterns and insights.

   1. Text Analysis: Sprinklr AI analyses text data, which can come from text entry questions and the "Other" option in multiple-choice questions. It breaks the text into meaningful phrases, each categorised with a sentiment label such as positive, neutral, or negative, along with an attribute. These attributes are generated based on the survey context and the responses received. Let's understand it with the help of an example.

      Example:

      1. Text input: Stores could be better organized. And there is no parking available in the store.

      2. Phases Detected:

         • Stores could be better organized - Negative sentiment

         • And there is no parking available in the store - Negative sentiment

      3. Attribute Detected:

         • Stores could be better organized - Facility Ambience

         • And there is no parking available in the store - Parking Logistics

   2. Descriptive Analysis: Descriptive analysis of survey data focuses on summarizing and presenting the results to provide a clear overview. It helps you understand what has happened by examining trends, patterns, and data distributions, without making predictions or testing hypotheses. Let's have a look at the tests that are run in descriptive analysis:

      1. Frequency Analysis: Measures the frequency of specific responses(Example: Percentage of "Very Satisfied" ratings).

      2. Measure of Central Tendency: Identifies the center of the data using the mean, median, or mode.

      3. Measures of Dispersion: Explores data spread with metrics like range and standard deviation.

         Let's have a look at the example for descriptive analysis:

         45% of respondents are aged 25-34, and 60% identify as female.

         The average satisfaction score is 4.2 out of 5, with 70% rating their experience as 'Excellent' or 'Very Good.'

   3. Correlation Analysis: Correlation analysis examines the relationship between two variables, where each question represents a variable and the responses are its values, to determine if and how they are related. In the context of survey data, it helps identify patterns or associations, like if a change in one response is linked to a change in another. Let's have a look at the broad catergories of correlation:

      1. Positive Correlation: As one variable increases, the other one also increases. Example: Higher satisfaction may lead to higher loyalty.

      2. Negative Correlation: As one variable increases, the other one decreases. Example: Longer wait times may lead to lower satisfaction.

      3. No Correlation: There is no identifiable relationship between the two variables.

         Example of correlation analysis

         • Higher customer satisfaction is strongly associated with more frequent shopping behavior.

         • Increased social media ad spend shows a positive correlation with higher website traffic.

   4. Regression Analysis: Regression analysis is a statistical method used to understand the relationship between one dependent variable and one or more independent variables. In the context of survey data, it not only identifies whether relationships exist, as in correlation analysis, but also quantifies the impact of the independent variables on the dependent variable and helps predict outcomes.

      1. Linear Regression: Analyzes how a continuous dependent variable is influenced by one or more independent variables.

         Example: How does satisfaction (1–5 scale) influence the likelihood of repeat purchases?

      2. Logistic Regression: Used when the dependent variable has two possible outcomes, such as yes or no, or satisfied versus unsatisfied.

         Example: What factors increase the likelihood of a customer recommending the product?

      3. Polynomial Regression: Examines non-linear relationships where data patterns do not follow a straight line.

         Example: How does spending behavior change as income increases beyond a certain threshold?

3. Process Results: Once the results are ready, Sprinklr AI+ generates a relevant heading for the report along with a subheading that summarizes the report. In the insights section, you will see 3 to 4 insight cards based on the analysis. Each card includes an action item, offering clear next steps to help improve performance.