Introduction

In simple terms, Insights Assistant is a tool that enables and assists in generating meaningful insights easily and in a timely manner. It leverages Sprinklr's AI capabilities to show the most crucial insights around key product metrics without requiring you to create dashboards or drill down on widgets. This will improve the efficiency and reduce the time it takes to identify insights.

Additionally, You will have the ability to save insights they identified. You can then conduct further actionability on insights and effectively collaborate on insights, allowing you to share insights across teams, assign them to others, and track progress on tasks, improving the cumulative efficiency of the team.

Benefits of Insights Assistant

Conventional Insights Collection Process:

In the past, you had to perform multiple steps for identifying insights and there was a need for multiple drill-downs to get to relevant information. Also, a number of times you are unaware of which widget to plot on the dashboard to identify insights and solve their use case. Moreover, collaboration is at the core of key business processes, and it is increasingly important for you to be able to effectively disseminate insights across their organization. Earlier, there was no methodology to save or share an insight that was identified by you. Apart from this, there was one more issue:

Quality of Insight captured earlier

Even after putting in efforts to generate an insight, you did not have visibility into why and how of the issue. Both of these issues and roadblocks have been resolved by using the Insights Assistant tool. 

Personas and benefits for Insights Assistant

• Product Managers can use the Insights Assistant to track product usage and identify areas for improvement. They can also use Boards to share insights with their team and collaborate on product strategies.

• Marketing Managers can use Insights Assistant to monitor the effectiveness of their campaigns by tracking key metrics and identifying trends.

• Data Analysts will find value with the AI-generated insights, which can help save time by providing a starting point for analysis.

These personas are not exhaustive, and the Insights Assistant tool helps a wide variety of people depending on their specific requirement and use-cases.

Insights Assistant Navigation

Insights Assistant Layout

The Insights Assistant consists primarily of five sections –

• All Insights: It contains both manually and AI generated insights displayed in separate tabs for a lucid segregation of both the insights.

• Created by me: This section is dedicated to store all the insights manually selected and added by you.

• Favorites: This section contains all the insights marked as favorite by you, by clicking on the star icon on the insights card.

• Archive: This section contains all the archived insights by you.

• Boards: These are the scrollable insights roster, which are strategically created to share the insights with your coworkers for a quick and collaborative brainstorming.

All Insights

The All Insights section is a consolidation of both Manual and AI-generated Insights. It consists of all the insights generated within Sprinklr displayed in the form of a stream/feed. It is bifurcated into two tabs – Manual Insights and AI Insights.

• Manual Insights: This tab will consist of all the manual insights which are created by all the users in Sprinklr. Laid out in a social media-esque style feed, it allows the you to scroll through every manual insight added and engage with multiple actions on them.

• AI Insights: This tab contains all the AI generated insights within Sprinklr. These insights are auto generated by the AI model and eliminates the requirement of manually selecting and adding every insight. The smart AI model detects the crucial dips, peaks, or anomalies which are most relevant to the product or brand.

The All Insights section also has 2 major filters:

1. Insights Type

   1. Trend: This type of insight indicates an increase/decrease in the metric concerned over a relatively long duration of time. This helps in identifying whether the entity concerned is doing well or poorly in recent times.

      ​

   2. Anomaly: This is like Trend Insight, but the major difference being about sudden drops or sudden increases in the respective metrics. An anomaly usually indicates drops that need instant attention.

      ​

   3. Low Performaing Product: Metrics can remain consistently low without showing significant spikes or drops, which limits insights to only when they increase or decrease. This misses about 40% of valuable insights for low-performing products.

      ​

      To address this, a new insight type identifies when a product or brand performs worse than the industry average. A product is considered "low-performing" if its average metric (e.g., star rating or experience score) is below the industry benchmark, based on vertical or category. Benchmark information will be provided before activating this insight for any client.

      Note: In the Anomaly and Trend cards, dates will be grouped into buckets. You previously saw an impact score for each card when drilling down into a card. This impact score has now been removed, and only the severity score is displayed upon drilling down. This score represents severity for negative cards, while for positive cards, it indicates relevancy.

      This update will average multiple severity scores using a weighted calculation. The resulting severity or relevancy score (depending on whether the card is negative or positive) will replace the impact score at the top of the card. You will also be able to sort the cards based on this new severity or relevancy scores.

      Additionally, you will have access to drill-down information and the root cause analysis.

2. Metrics of Insights Generation

   1. Star Rating: Review Star Rating from 0 to 5.

   2. Experience Score: A custom sentiment score is calculated based on the number of Insights from the mentions & reviews. Experience Score equals = 100* (Number of positive insights) + 50* (Number of neutral insights) + 0*(Number of negative insights) / Total Insights. Applicable Sources: All Listening Supported Channels (Earned & Owned).

   3. Custom Metric: You would be able to get insights on custom metrics and custom fields [number type] created on Sprinklr platform.

Created by me

Created by me consists primarily of the insights which are created manually by you. These insights will be added to the Created by me section of the app. You will also have an option to add the insights to the Board.

To add desired insight to the Insights Assistant –

1. Click the Options icon on the widget and select Create Insight.

   ​​

2. The create insight popup will have a snapshot of the widget as an Insight Preview. You will have the option to add the insight to the Created by me section and to add the insight to board. ​

   ​

3. The selected insight will now be added to the Created by me page.

   ​

Boards

Skimming through a huge volume of insights can be a daunting task and can overwhelm you with a lot of data which might not be relevant to their use case. Boards helps you to avoid the rigorous scrolling of data as it is a common place where the insights are shared only with the desired stakeholders. While creating an insight you can choose the option to create and add to the board, or create a new board.

You can then select the members with whom they want to share the board.

The board will be created in which the you can add more members and sections later as per their requirement. The board also lets you add AI-generated Insights with the option of AI Insights Setup.

Here you will have the option to automate the AI insights generation based on the filter criteria. You can set up multiple filters and define their values to fine tune and reduce the time to insights.

Additional details on Sources

Insight Assistant is a powerful tool that leverages review data and metrics to provide valuable insights to our clients. It has been successful in analyzing Experience Score and Star Rating, which has garnered positive feedback from you. Additional data sources are available, including social data from topics/themes, FPDI, and Accounts.

AI Insights

How does it work?

The enhanced Insights Assistant combines our inbuilt generative AI model with data analysis, severity assessment, and metric averages. This synergy allows the AI to identify anomalies, trends, and areas of improvement both from brand and product-centric perspectives. The ultimate goal is to empower you with smart recommendations that contribute to improving your metrics and you experience.

Key benefits

• AI-generated Recommendations: The AI generates actionable recommendations that are displayed on insight cards and can be further explored through detailed drill-down views. This eliminates the need for manual investigation, enabling you to quickly identify and address issues related to low-performing attributes.

• Impact Metrics: For each recommendation, corresponding impact metrics are presented on the user interface. These metrics include Severity/Relevancy scores, % contribution, and average metric values. This contextual information helps you understand the importance of each recommendation in relation to your metrics.

• Improved Prioritization: The AI-generated recommendations are prioritized using impact metrics. This prioritization guides you directly to the most impactful recommendations that demand immediate attention, enabling efficient decision-making and action-taking.