An engagement on TRAC is a reaction to a post, for example a repost, a share, a reply or a comment, and this varies depending on the platform or data source.

Before we delve into the types of engagements we collect, firstly we need to distinguish how we view and store these engagements on TRAC.

The first type of engagement is what is stored as an actual piece of content, i.e. you can perform a full text search on this in your results, and this is true for reposts, replies or actual comments. So on TRAC, reposts and comments are treated as a type of post, and therefore when we sum up the volume of data collected in your search, and over time, these two types of engagements are counted towards your overall volume of mentions. In other words, the total volume of mentions in a search = Posts + Engagements, where a post is a tweet or a blog post, and an engagement is a repost, a comment, or a reply.

​

The second type of engagement on TRAC is an engagement as a count. This type of engagement data is stored simply as a number, because it is not a physical piece of content with text that you can read or search on. Examples of this are the number of shares or number of comments, or number of likes. Because these counts are just a number, instead of a physical piece of content with text, these types of engagements are not factored into the overall volume of data collected in your search. Hence they will not be added to the total volume of data displayed in the Key Numbers charts, nor in the Overview Mentions graph. The reason for this behaviour is down to the way that Pulsar indexes different types of data and we wouldn't want comment_counts, like_counts and share_count to count towards your data volumes, as that would be huge!

Now that we've explained the 2 types of engagements that we collect, below is a breakdown of the types of engagements that we collect for each data source on TRAC.

We collect reposts, replies, and no. of likes. However, as noted above, the number of likes is stored separately and not summed up to the total engagements in the Key Numbers chart or the Timeline charts. Only reposts and replies are factored into this metric, and therefore will count towards the overall volume of data collected in a search.

NB: We collect matching replies for historic data (replies that contain the key terms), and all replies for real-time data

In a Facebook Page Panel search, we collect the following:

As noted previously, the no. of likes and the no.of reactions are stored separately, and not summed up to the total engagements in the Key Numbers chart or the Timeline charts. Only the actual comments are factored into the volume metric, and therefore count towards the overall volume of data collected in a search.
​

In a Facebook Topic search, the Facebook engagements we collect are the no. of likes. And if available, we also collect the number of comments. We are unable to collect the actual comments due to recent changes enforced by Facebook. And as mentioned previously, the no.of likes are not factored into the Key Numbers chart or the Timeline charts, and therefore are not counted towards the overall volume of data collected in your search.

Engagements on these channels are the no.of social shares that the article has received on Facebook. In other words, this is the number of times the article URL has been shared on Facebook. For Reviews, we also also get the Review Rating that has been left by users on the site. And similar to the no. of likes and no. of comments, the no. of social shares is not counted towards your overall volume of mentions.

Following Instagram API changes, we only collect the no. of likes and the no. of comments, not the actual comments that a post has received. So, for Instagram, as previously explained at the beginning, when looking at the total engagements from Instagram in the Key Numbers chart, this will always show 0, unless you switch to the Results view, as shown below, where you will be able to see the like count and comment count for each post.

The only exception to this rule is if you are using the TRAC functionality which allows you to collect tagged mentions of your brand. This is a common practice on Instagram, where a user tags your business or brand in a photo or video. In this case we are able to collect the actual comments, in addition to the no.of comments and no. of likes.

For Reddit, we get the number of shares, which count as engagements. Replies to Posts are collected as separate posts themselves. We also get the number of upvotes (likes) for every Reddit post.

For Radio we don't collect engagements. We collect Media Reach, Duration and AVE.

For TV we don't collect engagements. However we have the Media Reach, Duration and AVE.

For Podcasts we don't collect engagements. However we have the Media Reach, Duration and AVE.

For Print data we don't have engagements, however we collect the Circulation figure as well as AVE.

For Threads, we collect the number of Likes and number of Replies for engagements. However, we only ingest replies that match the search query, similar to Youtube.

We only collect the video views for Twitch.

For Pinterest we collect the number of repins or saves and the number of comments left on a pin.

For Discord, engagements are only collected on Thread type posts. Otherwise, all comments are considered posts in their own right.
​
Threads are a unique type of conversation on discord that collect all replies within a singular grouping rather than in-line with the rest of conversation as most of discord functions. It's not an incredibly common conversational tool within discord servers but it depends on the server!

YouTube provides an engagement count based on the total number of comments on a video, and also the actual comments that match your topic search As such, the total number of comments and the actual comments may differ, and that's because the actual comments will be fewer, since we are only indexing comments that only contain your keywords.

For VK we collect the number of number of comments only. We do not collect any other type of engagement metric or likes.