Mention Classifications and Widget Metrics Glossary for Cision Social Listening

Modified on: Tue, 22 Nov, 2022 at 7:04 PM

The Mentions Panel:

1. The right side of the dashboard on the Analyse page includes a collapsible mentions panel. You can sort posts in the mentions panel by different metrics, depending on whether you are looking at a Quick Search or Saved Search.

For Quick Searches, the sorting options include Newest First, Oldest First, Random, Most Relevant and Least Relevant.

For Saved Searches, the sorting options include: Highest Reach, Newest Post First, Oldest Post First, Random Order of Posts, Highest Twitter Retweets or by Highest Twitter Impressions (potential number of times a tweet may have been seen).

2. Click on the mention itself to see the Mention Details, including several metrics associated with the mention. 

3. Click the Original Post icon to be taken to the original post. 

4. The date and time of the post will be displayed, along with metrics that can be edited, including Location, Language, Sentiment, and Emotion.

5. You may also click on the Add tag button to tag the mention.

Click on the trash can icon to delete the mention from your results.

Click on the arrows at the bottom of the box to move to the Mention details for the next mention.

Click on the X in the upper-right-hand corner to exit the Mention details.

6. Click on the Show More link to expand the Mentions Panel to view all mentions.

7. Click the Hide link to remove all Mentions and the Mentions panel from the dashboard.

8. Click on the Show Less link to return the Mentions Panel to the right side of the screen.

9. Click on the Export link to download the entire list of mentions into a PDF or .CSV file.

NOTE: Clicking into areas within the dashboard automatically filters the mentions on the right to show the specific mentions behind a given insight.

Exporting Mentions into a PDF:

To export mentions into a PDF document:

1. Click on the Export link in the upper-right corner of the Mentions panel.
2. Make selections for your export:
  • Select the PDF format
  • Select the number of mentions you would like to export. NOTE: The maximum amount of mentions you can export into a PDF is 500.
  • Click on the EXPORT button.

The exported mentions will display exactly like in Social Listening, including images.

Total Volume Widget:

Get a quick overview of top-level numbers to see how your saved search has changed. Use this to quickly monitor brand reputation and campaign performance. Metrics 1-4 are the default metric for this widget.

1. Total Mentions: The total volume of social posts that match your search criteria in the period selected and filters applied.

2. Total Reach: Sum of all reach scores for all individual mentions in your search, period, and filters. Reach is designed to estimate how many different individuals are likely to have seen that content.

3. Total Impressions: Twitter's only metric that estimates the potential number of times the tweets in your search may have been seen calculated by adding author and retweet followers in your data set.

4. Unique Authors: The number of unique authors in your data. Compare this to the number of total mentions for insight into how active and engaged your audience is. The greater difference between mentions and this number shows the audience is more engaged and posting more often about your search terms.

5. Retweet Rate: The average number of retweets per tweet a search receives. This metric lets you know how viral a topic is (i.e., if the rate is high, then it shows a lot of people were in agreement with or supported what was being tweeted) or how much organic conversation occurred around a topic (i.e., if the rate is high then not a lot of organic conversation occurred around a topic).

6. Average Followers: The average number of followers of all contributors. This metric explains if a specific topic has high clout in general. For example, if the average follower number is lower than it was previously, then we know that people who are talking about this topic are potentially less influential in general. A higher average follower number would indicate that people involved in the conversation have more followers and thus have the potential to be more influential.

Percent Change: Takes a look at the data range you've selected and compares it to the same previous date range. If it’s set to look at the past 1-30 days, the Total Volume Widget will display the percentage change and value for the previous 31-60 days.

To select which metrics you would like to display in the Total Volume widget:

1. Click on the Cogwheel icon in the Total Volume widget.

2. Using the dropdown on the right-hand side of the screen for each metric, select up to 6 metrics you would like to include. Select "None" from the dropdown if you do not want a metric displayed.

3. Click on the X button to close the configure tile screen.

NOTE: The widget will reset to the default metrics 1-4 when you leave the page. (Total Mentions, Total Reach, Total Impressions, Unique Authors).

Mention Volume Over Time Widget:

Shows the number of mentions over time with options to look over an hour, day, week, or month. This can be used to see how your brand performs over time, how a campaign may impact discussion, or even how a crisis is developing. 

Mentions: the volume of social posts that match your search criteria 

For Saved Searches, drop-down options allow you to see the number of mentions by the hour, day, week, or month. 


For Saved Searches, AI-powered peak detection surfaces significant shifts in data. Clicking into a peak shows an explanation of the drivers for mentions that go viral.

Mention Volume by Day of the Week and Hour Widget:

Understand when the subject of your search is most popular to impact your content calendar. 

Mention Volume by Day of the Week and Hour: This shows a heat map of the days of the week and time of day that conversation within your search is most popular and getting the most volume of mentions.

Mention Volume Benchmarks

Compare volumes from a selected date range against the volumes of a preceding date range of equal length.

Mention Volume Benchmarks Displays data from a selected period while comparing it to another period of the same length. You can compare data for the Platform, Sentiment, Emotion, Interests, Country, or Language using the dropdown.

The percentage displayed lets you know if there was an increase or decrease from the previous period.

The horizontal line will display the average of the data for the previous period.

Ex: Compare June-July to April-May.

Note: Year-over-year data is not available at this time.

Sentiment Volume Widget:

Understand the view consumers have around your search query – whether it’s positive or negative.

Sentiment Volume: This shows the number of mentions classified as negative, positive, or neutral as a breakdown of the total mentions.

The key will display the percentage of total mentions for positive, negative, and neutral sentiments.

When hovering over a sentiment type in the donut chart, you will see the percentage and the total number of mentions.

The sentiment is assigned to Mentions using an in-house model developed by Brandwatch’s data scientists. The team compiled a collection of ~500K documents and labeled them as positive, negative, or neutral. These hand-labeled mentions are then used to calculate the frequency of distribution of each word, negated word, emoticon, etc., present in those mentions across the positive, negative, and neutral categories. These frequency distributions are then used to construct a model that classifies each new mention.

Sentiment Volume Over Time Widget:

Shows the number of mentions with positive, negative, or neutral sentiment over time with options to look over an hour, day, week, or month. This can be used to see how your brand performs over time, how a campaign may impact discussion, or even how a crisis is developing. 

The Sentiment Volume Over Time uses a line chart to display the volume of mentions over the dedicated period for each sentiment category. 

  • For Quick Search results, the volume of mentions is displayed by days as data only dates back 30 days. 
  • For Saved Searches, drop-down options allow you to see the number of mentions by the hour, day, week, or month. 
  • Like mentioned volume, sentiment also has peak detection that uses AI to surface significant changes in data for Saved Searches. Clicking into a peak shows an explanation of the drivers.

Emotion Volume Widget:

Understand people's feelings about your brand by seeing how they talk about it in their social content. This can be extremely valuable to measure during and after a crisis to determine overall brand reputation impact. 

Emotion Volume: Bar chart showing the number of mentions coded with certain emotions.
Emotion is assigned to Mentions using an in-house model developed by Brandwatch’s data scientists. The data scientists compiled a collection of ~2M documents labeled with emotional categories and trained a model to predict the emotion based on the features of the document.  Each document in the collection originally ended in an “emotional” hashtag, which we then removed and trained the model to predict the emotion for each document based on the remaining text.

Emotion Volume Over Time Widget:

Track how consumers feel about your brand, company or product over time. Use this to see how campaigns can change the way consumers feel about your brand or measure response to a crisis over time. 

Emotion Volume over time: Shows the total number of mentions for each emotion for the selected time period. 

  •  For Saved Searches, drop down options allow the user to see the number of mentions by hour, day, week, or month. 

Word Cloud Widget:

Get a quick overview of what is being talked about within your search. This is a great way to understand the conversation around your brand, competitors, industry or campaign for overall brand reputation management, to understand campaign influence and resonance on consumers and what is happening quickly during a crisis. 

Word Cloud: contains the most frequently used terms and phrases within the mentions for a search query. The bigger the word, the more often it is used compared to the other terms.

Tip: Dive into different types of terms and phrases used and what the word size represents in the configure tile to learn even more. 

 Click here to learn how to Customize the Word Cloud for Saved Searches.

Top Shared URLS Widget:

Understand the articles, and content people are sharing around your search. This can help identify new outlets or influencers to target and whether coverage resonates with your audience.

Top Shared URLs: List of the top 10 shared links on Twitter, sorted by the frequency they appear in mentions. 
  • For Saved Searches: Up to 50 results will display. Click on the arrows at the bottom to see the next list of 10 Top Shared URLs.

Top Sites Widget:

See where conversations are happening about your brand. 

Top Sites: List the top 10 domains in the results where mention occurs, sorted by frequency of mentions. 
  • For Saved Searches: Up to 50 results will display. Click on the below arrows to see the next list of 10 Top Sites.

Top Authors Widget:

See who is talking about your brand.

Top Authors: List of the top 10 authors across any platform for which there is a populated author field (usually Twitter, Reddit, and Tumblr), sorted by frequency of mentions. 

Use this widget to access influencer metrics, such as Reach, Follower Count, and Location.

Sort Top Authors by Reach, Twitter Follower Count, or Total Number of Mentions.

  • For Saved Searches: Up to 50 results will display. Click on the below arrows to see the next list of 10 Top Authors.

Top Interests Widget:

Get an understanding of your brand’s end customer talking about you. Use this to formulate campaigns and messaging or see if you’re impacting the right people. 

Top Interests: List the top 10 interests associated with the authors of the mention returned by the search. These are generated from Twitter biodata, and the Top Interests widget will list the top 10 of the 22 interest categories available on Twitter. These are then sorted by mention frequency.

The 22 interests categories on Twitter include the following: 

Animals & Pets, Automotive, Beauty/Health & Fitness, Books, Business, Environment, Family & Parenting, Fashion, Fine Arts, Food & Drinks, Games, Home & Garden, Movies, Music, Photo & Video, Politics Science, Shopping, Sports, Technology, Travel, TV.

Top Hashtags Widget:

Understand the conversation around your brand by seeing the hashtags associated with it. 

Top Hashtags: List of the top 10 most used hashtags in the search’s Twitter mentions, sorted by frequency. Select a hashtag to see a list of all the mentions that use it. 
  • For Saved Searches: Up to 50 results will display. Click on the below arrows to see the next list of 10 Top Hashtags.

Male vs. Female Breakdown Widget:

Get an understanding of who is talking about your brand. 

Male vs. Female Breakdown: Depicts Twitter data with identifiable male and female names to provide a comparison of who is talking about your brand.

How Brandwatch Obtains Male/Female Data: 

This metric is calculated by matching the first name in an author’s Twitter profile to a curated dictionary of almost 50,000 names. If an author has a name that falls into both categories or one that does not match the dictionary, the male or female category will not be assigned to that author. While it is not a perfect methodology and is not inclusive of all gender identities, it helps you to better understand audiences and optimise strategies. 

Top Languages Widget:

Understand who is talking about your brand by understanding their language.

Top Languages: A list of the top languages found in the query results, sorted by the number of mentions and the number of unique authors found writing in each language.


Volume by Countries Widget:

Understand who is talking about your brand by understanding where they are located.

Top Countries: A list of the top 10 countries in the search results, sorted by mention frequency and listing the number of unique authors found from each country. This also is displayed on a map, with darker colours indicating more frequency.  

Some location data is provided directly by the platform and or the author of the Mention. When an author or a Mention has no explicit location, we attempt to infer a location based on the keywords in the author’s bio. We use a statistical classifier trained on the author bios of authors who’ve explicitly shared their location. This classifier uses features it has identified through training that implicitly indicate the location (e.g., the name of a local sports team, a city nickname, etc.).

Topic Wheel Widget:

Get greater context around how the most frequent topics associated with your search are being used for a more complete picture of how to adjust or clarify messaging.

Topic Wheel: Shows the most popular topics associated with your search in the inner circle and the associated subtopics in the outer circle. 

NOTE:  It is only available for Saved Searches. 

Content Source & Measure of Engagement Metrics


  • likes, comments 


  • reactions, comments, shares 


  • retweets, replies, followers of retweeters, followers of the author 

Others (blogs, forums, news, etc.) 

  • Average site visitors, average engagement, Alexa monthly visitors 

More on reach: 

  • Reach is heavily influenced by post engagement and traffic for the author or site. Our proprietary algorithm then uses each available metric and applies various (content source specific) assumptions based on previously observed behavior to infer how they may translate into the number of individuals likely to have seen a given post. It is not necessarily true that everyone who follows you sees your post or that everyone who follows you, so adjustments must be made to account for this and factor in the degree to which increased post engagement influences the final reach value. 

  • To improve your understanding of the processes involved, please consider the following example, in this case, using Twitter as the content source. 

    • Suppose a piece of content is shared on Twitter by an author with 5000 followers. It has so far gained 3 likes and 2 replies and has been retweeted by 1 person with 3500 followers. We must now estimate the number of people, X, likely to have seen this content. 

    • This number, X, will be significantly lower than the sum of poster and sharer followers because we know that posts with this kind of engagement won’t be prioritised in all followers’ feeds. We can also be confident that X will be higher than for a post from the same author that doesn’t have any engagement, as we also know that any engagement will lead to a higher priority on followers’ feeds. Visibility will be further increased by the extra exposure that this engagement inevitably produces. For example, a particular account that doesn’t follow either the original tweeter or the retweeter may follow the “liker” and gain exposure to the post. 

    • We, therefore, need to estimate how many, on average, of the tweeter's followers normally see a post and consider how this increases if this post has a like or a comment and how this changes if a piece of content is shared. We then consider the observed metrics of the post in question (3 likes, 2 comments, 1 retweet) and apply some scaling to reflect this relative engagement. Our algorithm assigns a value to this and adjusts the reach accordingly. It also considers how the different sites prioritise content on their platforms. 

    • Note that a reach of zero is not unlikely for a post by an author with a low number of followers; reach is an estimate not an exact count.  

  • The algorithm is not overly reliant on follower count alone. For example, it’s quite reasonable to assume that a tweet by an author with only 30 followers has not been seen by anyone. When a tweet is retweeted or replied to, its reach increases, particularly if the retweeters/commenters have high follower counts. 

Was this article helpful?

That’s Great!

Thank you for your feedback

Sorry! We couldn't be helpful

Thank you for your feedback

Let us know how can we improve this article!

Select atleast one of the reasons

Feedback sent

We appreciate your effort and will try to fix the article