Listening Insights

What is Sprinklr Listening Insights?

Sprinklr’s Listening Insights is a set of features that enables users to explore publicly available social media data collected by Sprinklr through specific search queries built around keywords. All messages matching these queries are brought into the system, allowing measurement, monitoring and reporting at the message and at an aggregate level.

Listening Insights is composed of topics, themes and keyword lists. Messages are reviewed on listening dashboards or listening columns placed on engagement dashboards. Read below for more information on these tools, their use cases and instructions.

Listening Insights Access Restrictions

Listening Insights is an added functionality with an additional cost to MSU. This functionality is paid for centrally by University Communications and Marketing. MSU has an annual allotment of messages to monitor within Listening Insights, and going over this allotment may result in substantial additional cost.

For this reason, University Communications and Marketing restricts access to Listening Insights functionality to limited strategic needs, either driven by University Communications and Marketing alone or in partnership with select units. If the unit has a use case for listening, University Communications and Marketing is happy to discuss it and to determine the best means for meeting the unit need. This may involve Quick Search, Listening Insights or another means.

Feature Information

There are a few limitations that users may bump up against when working with Listening Insights. Below are a few of the more commonly experienced limitations:

Maximum number of topics in a topic group: 100
Maximum number of keywords in a keyword list: 200
Maximum number of widgets in a dashboard: 25
Maximum number of rows in an Excel widget data export: 10,000
Maximum proximity of terms when creating a topic: 20
Wildcard limit for a topic: 1
Maximum number of keywords in a topic query: 1,000

To read more about Listening Insights platform limitations, please see the Sprinklr Knowledge Portal.

Listening Insights will pull in publicly posted messages based on the topic query and system settings and application programming interface (API) limitations. Certain parameters, such as the blocklist, prevent messages from entering the system for any topic.

The following platforms support full content search, as of March 8, 2021: Facebook Reviews, Flickr, Glassdoor, Google Reviews, Quora, Reddit, TripAdvisor, Tumblr, X (formerly Twitter), vKontakte (VK), WordPress, YouTube, blogs/websites and comments, forums, news, TV and radio, print, videos and comments, reviews, and Yelp.

The following platforms only support search within owned content, as of March 8, 2021: Foursquare, Line, LinkedIn, Pinterest and SlideShare.

The following platforms have limited search, as of March 8, 2021: Instagram (hashtags on parent posts only, no comment search, no private account search, @ replies using hashtags are tracked via themes but not topic queries and hashtag search volume limitations are in place) and Weibo (formerly Sina Weibo) requires additional capability enablement.

To read more about Listening Insights source restrictions, please see the Sprinklr Knowledge Portal articles on listening sources.

To request access to Sprinklr Listening Insights, contact University Communications and Marketing at Comms.WebSupport@msu.edu. University Communications and Marketing cannot guarantee requests will be approved, but staff members are happy to discuss use cases.

Listening Insights Terminology

Basic Settings

Keyword lists — Using keyword lists allows for re-use of the same set of words across multiple topics and only need to update the list in one spot. University Communications and Marketing recommends building somewhat granular keyword lists that can be combined in different ways to avoid hitting the maximum keywords in a topic limit. For example, rather than a “sports” keyword list, we’d suggest sport-specific lists (i.e., basketball, football, soccer) that can be combined as appropriate. Read more about using keyword lists in the Sprinklr Knowledge Portal article, “Introduction to Keyword Lists.”
Themes — Themes are sets of custom filters and keywords used to sort data. Read more about using themes in the Sprinklr Knowledge Portal article, “Introduction to Listening Themes.”
Topics — Topics are the main component of listening. They are built using Boolean queries (see Listening Basics) and various rules, and volume limits and time parameters can be added to control how many messages the topics bring into the system. They can be grouped into topic groups for ease of organization and reporting.
Topic backfill requests — A topic backfill will draw in data on a topic from before the topic’s date of creation. The system does not allow users to control how much volume the backfill will bring in so it can put MSU much closer to the consumption limits without any ability to stop it. For this reason, requests are limited to those with admin level Listening Insights access and are rarely, if ever, used.

Advanced Settings

Using Listening Insights

Types of Listening Insights Topics

There are five types of queries in Sprinklr Listening Insights:

Query-based listening — This uses a Boolean search approach to listen to messages from social media and web sources.
Account-based listening — Listen to messages, comments and posts on paid and owned accounts.
Profile/Domain-based listening — Listen to specific profiles and web domains.
Location-based listening — Listen to user content from a specific area or location.
Weibo-based listening — Listen to messages and comments on Weibo.

Create a Listening Insights Topic

Listening Alerts

Sprinklr offers two categories of alerts: smart alerts, which are powered by artificial intelligence, and custom alerts, which are programmed using manual rules. Utilize alerts to send notifications (push, email and/or system) when listening picks up trends that match the rules (for manual alerts) or that set off the AI-based sensors based on the type of smart alert. For example, listening alerts can notify staff of unusual activity on a listening topic. Staff can go about their day without constantly monitoring the system, knowing that if there is something that requires attention Sprinklr will send a notification.

Create Smart Alerts

Open a New Page tab in Sprinklr.
Navigate to the Platform Modules menu section and click “Alert Manager” under the Mitigate Brand Crisis sub-menu in the Love column. If a user does not see this option, they do not have role permissions for alerts. Contact University Communications and Marketing to request access.
Click the New Alert button (top right corner).
Name the alert. Be sure it is specific enough that it is clear what the alert is doing. Users may also add a description and tags to assist with sorting/filtering alerts and to help others know more about the alert’s intent.
Be sure “Listening” is selected from the Module dropdown.
Choose “Smart Alert” under “Category.”
Select the main filter dimension(s). Multiple dimensions can be selected. The options include:
1. Language
2. Country
3. Topic Group
4. Sentiment
5. Topic
6. Topic Tag
7. Theme
8. Theme Tag
Select the additional dimensions required based on the selection in step No. 7.
Choose the Smart Alert Type(s) from the dropdown.
Quick tip: Users can choose multiple types in one rule, but they will all contain the same subject line for emails. To differentiate alerts in the alert message, create a different alert rule for each alert type and customize the message.
Add additional filters to refine the alert conditions, if necessary. Options include:
•    Mentions
•    Alert Severity
•    Alert Longevity
•    Alert Severity Increase
•    Total Reposts
•    Reposts Gained
•    Alert Type
Select the alert actions. Options include:
•    Send Email Notification — Listed recipients (Sprinklr users, user group members or external email addresses) receive notification via email
•    Mobile Push Notification — Listed Sprinklr users receive a Sprinklr app push notification. The user must have the app installed on their phone and be logged in to their Sprinklr account.
•    Exclude Reposts in Email
Adjust settings for the alert, such as subject line, from address and timestamp messaging.
Optional: Check “Add to Queue” to add the alert to an alert queue. The alert queue must be created first.
Click “Save.”
Monitor the alert and adjust settings as necessary.

Note: Smart alerts can take up to two weeks to “learn” what behavior is average for a topic/topic group before the alert will be able to report unusual activity. If an issue is urgent and new, consider supplementing the smart alert with a manual alert.

Create Custom Volumetric Alerts

Documentation updated: Nov. 1, 2024

Annotations — Annotations allow us to add notes regarding the data, such as noting a significant event that drove a spike in conversation volumes.
Blocklist — The blocklist can include user profile lists, sources, keyword lists and individual keywords that will be excluded from any listening query in the system. These don’t count toward the keyword per topic limit. University Communications and Marketing commonly uses these to block out accounts and messages that are false positive matches to queries, such as other MSU institutions (i.e., Mississippi State or Montana State) and spam accounts. Read more about the block list in the Sprinklr Knowledge Portal article, “Listening Blocklist.”
Content links — Lists of content links can be used to filter and route messages within Listening Insights. Any message containing the link in the body of the message can be routed using these rules. Read more in the Sprinklr Knowledge Portal article, “Listening Content Links.”
Custom metrics — Should the need arise, the system does allow for creating custom metrics in Listening Insights. Read more about what metrics are already available in the Sprinklr Knowledge Portal article, “Metrics and Dimensions Supported in Listening.”
Domain list — Domain lists are used to create lists of domain addresses that can then be used in topic builds to either include or exclude the domain as a source. Read more in the Sprinklr Knowledge Portal article, “Listening Domain Lists.”
Generating sources — Individual sources of data can be turned on or off at the topic level. For example, users may choose to build a topic that only looks at Facebook and X. University Communications and Marketing typically leaves TV, radio and print off for topics because the department uses additional tools to monitor brand references on these channels.
Entities — Entities are lists of famous people, things, brands or products. Users can use listening to measure message counts against these lists.
Explore — The Explore functionality allows users to apply a dashboard-based or widget-based overlay to Listening Insights data to drill across multiple dimensions simultaneously. Read more in the Sprinklr Knowledge Portal article, "Explore Lens."
Target metrics — The system allows for setting targets for various metrics to measure performance and determine if goals have been achieved.
Topic templates — Topic templates can be used to structure the ways topics are created or apply rules to existing topics. Read more in the Sprinklr Knowledge Portal article, “Use Topic Template to Create a Topic.”
Scheduled exports — Exports of raw data can be scheduled. This may be useful for feeding data into external databases.
Scheduled reports — Reports can be scheduled and delivered via a variety of forms and channels to Sprinklr users and nonusers. These reports are based off dashboards (either standard dashboards or custom dashboards).

Listening Insights