Listening Topic Insights
After you create a Topic in Sprout Social Listening, it gathers data so you can observe trends, uncover patterns and gauge emotional response with regards to particular keywords, hashtags, brands or industries.
You can view Topic Insights for any Active or Archived Topic. Archived Topic data is only available for the time period that Topic was active.
This article contains the following sections:
Customizing Topic Insights
To view the insights that your Topic query pulls in, click the name of the Topic from the Listening screen in Sprout. You can also see a list of your active Topics in the Left Bar on the screen.
The first time you go to review insights for a Topic, it can feel like there’s a lot to look at. To make the data more manageable and meaningful, start by selecting the date range you want to view insights for. You can select a custom date range for comparison by clicking on your desired dates.
If you want more details, click and then click About these Insights to get a glossary with definitions, tips and tricks about your insights.
Use filter groups from the filter menu to further customize your insights. When you select filters, the message data you see throughout the Topic Insights (Performance, Conversation, Messages and Themes) gets updated based on the filters you select.
In the Profiles group, you can select which network and competitors you want to view insights from. If no network or competitors are selected, you see insights from all networks and competitors in the main screen. Select a particular network and/or competitor to only see those insights.
Use the Theme group to see only messages that relate to a certain Listening Theme. If no Themes are selected, you see insights from all Themes in the main screen. Select a particular Theme to only see those insights.
Use Keyword filters to include or exclude keywords in the message insights. For instance, if you have a Brand Health Topic for your coffee company, but only want to review insights for messages that contain your brand name, enter your brand name in this box.
Apply Advanced filter groups to narrow your insights even more by focusing on Content Type, Message Type, Sentiment, Technology or Mobile. Additionally, as you navigate through your Topic Insights, you can select data in the four sections that automatically apply filters to the insights.
If you used Themes on your Topic, you can also filter by those Themes. Select All Themes or drill into messages related to one particular Theme. To learn more about applying Themes to your Topics, review the Listening Topic Themes article.
Navigating Topic Insights
Topic Insights are separated into five sections:
This section of your Topic Insights starts with a summary of the overall Topic, or a summary based on the filters you apply. This section gives a quick overview of volume statistics, total engagements, potential impressions, unique authors and sentiment.
The Volume chart shows changes in your message volume to help you spot conversation trends. You can select how you want your message volume broken down using the dropdown. Click on a particular day in the chart to view the messages from that day.
The Potential Impressions chart helps you analyze how many potential impressions the Topic received. You can update this chart by breakdown or comparison. Click on a particular day in the chart to view the messages from that day.
The Engagement chart helps you analyze user engagement to see how audiences interact with the content related to your Topic. Choose from overall Engagements or Comments, Shares and Likes and choose the breakdown or comparison you’d like to see. Click on a particular day in the chart to view the messages from that day.
The Sentiment Summary chart is a snapshot that comes from positive and negative messages from the time period you selected. Sentiment trends help you fresh things out by seeing the changes in human conversation over time. It can be helpful to dig into these messages to find out what’s going on if you notice a large change in sentiment.
The Sentiment Trends chart shows how sentiment changes over the time period you selected. If you notice something isn’t quite right with the way the sentiment is classified, you can change the associated sentiment.
The Network Overview chart gives you key metrics for each network you selected for the Topic. You can sort this table by clicking the column names and update the organization by clicking and selecting which metrics you want to appear in the table.
The Conversation section of your Topic Insights provides more details about the keywords and hashtags used most often in the messages your Topic pulls in.
The World Cloud shows the top keywords, hashtags, mentions and emoticons found in the Topic. The words come from the messages that audiences published. People say more in their social media messages than just the keywords you put in your query. You can click on the keywords, hashtags, mentions and emoticons in the Word Cloud to go a little deeper into the messages. Then, when you go back to the Performance tab, everything in that tab is updated based on the word you selected in the Word Cloud.
You can then use the terms in the Word Cloud to help update your query.
The Related Keywords & Hashtags table helps you analyze the most common keywords and hashtags in your Topic. You can see a breakdown by sentiment for related keywords and hashtags.
The Conversation Overview table provides key metrics for keyword and hashtag in the Topic. You can sort this table by clicking the column names and update the organization by clicking and selecting which metrics you want to appear in the table.
The Smart Categories table shows you key metrics for the people, places and things discussed in your Topic.
Note: Smart Categories are only available in English.
The widget uses machine learning to identify entities and classify them into categories. These categories are:
- Person: includes people and fictional characters.
- Place: includes locations, buildings, airports, geopolitical areas and other locations.
- Product: includes objects, vehicles, food and other products.
- Organization: companies, agencies, institutions and other organizations.
- Media: includes books, movies, works of art and other media.
- Event: includes sporting events, concerts, natural disasters, wars and other events.
- Demographics: includes languages, religious or political groups, nationalities and other demographic factors.
- Policy: includes laws and legal documents.
Note: If the model isn’t confident in a Category, it won’t tag the Category to avoid false positives. This can happen if a message text is short and there isn’t enough context to discern the Category. A message may also be tagged with multiple categories if it includes more than one person, place and/or thing.
You can sort by clicking the column names and update the table by clicking and selecting which metrics you want to appear in the table.
You can also view example messages, include and exclude terms and compose posts with keywords.
The Demographics section of your Topic Insights provides more detail about the people participating in conversations about your Topic.
The Device Demographics chart shows a breakdown of mobile and web and Android and iOS that authors use when contributing to conversations.
The Profile Overview table shows metrics for specific profiles, including their followers, volume and impressions. This way, you can see who your main influencers are and find opportunities to connect with them.
The Locations map shows where the most conversations are happening about your Topic. For a better breakdown, the Locations Overview provides metrics by Country, Province/State or City. You can sort this table by clicking the column names and update the organization by clicking and selecting which metrics you want to appear in the table.
If you apply Listening Themes to your Topic the Themes section of your Topic Insights compares key metrics relating to those Themes.
You get a breakdown of each Theme, the Theme Group it belongs to, along with message volume, Share of Voice and other data around the Theme. You can sort this table by clicking the column names and update the organization by clicking and selecting which metrics you want to appear in the table.
The Messages section of your Topic Insights provides the raw data about the messages in your Topic. This is where you can find the raw data, the message in their entirety and the author that published it. You can see what they said sorted by date, followers or engagement.
To learn more about message actions in Listening, see this Help Center article.
Uncovering deeper Topic Insights
You can use data from the messages that appear in Insights to uncover trends in messages returned from your Listening Query.
Hover over charts to view all messages associated with that metric or chart.
Highlight a word or phrase (up to three words) in a message to view all messages containing that word or phrase, include the word or phrase as a filter, exclude the word or phrase as a filter or exclude the word or phrase from your Topic Query.
Click a message author name to view messages from that author or include or exclude that author as a filter.
Topic Insights FAQ
How much historical data do I receive for a new Topic?
30 days of data is pulled automatically for all networks except Instagram. This should give you enough data to decide if your Topic is set up properly.
How far can data be backfilled?
Twitter: All Time
YouTube, Tumblr: Sampling all time, but less data as time goes on
Reddit: 12 Months or 1200 Posts
Web: 25 Months
How often is data collected?
Twitter: New tweets brought in within one to two minutes of creation.
Facebook: Checked once every four hours.
Instagram, YouTube, Tumblr, Reddit: Checked once every four hours.
Web: Checked once every 24 hours.
Why isn’t there engagement data for pulled tweets?
Engagement data is updated for tweets anytime there is a retweet. If we don't receive a retweet, or replies and likes happen after the last retweet, we won't have the most recent number.
Why are some messages or comments not picked up by my Topic even though they contain the keyword?
It depends on the network. Twitter is the only network where we are confident we will receive every message. Other sources are sample based, which is a limitation of the networks.
Why aren’t some Instagram messages being pulled when the Topic keyword is in the caption?
We find Instagram messages by hashtag. Even if the keyword appears in the post, it may not contain the hashtag meaning Instagram won’t send us the data.
Why don't I have location data for all networks?
Not every network supports location data. Only Twitter and Instagram provide location data to apps like Sprout.