All Collections
What's New?
July 2023
Understand Public Discourse with Narratives
Understand Public Discourse with Narratives

Introducing Narratives analysis on TRAC

Linda Maruta avatar
Written by Linda Maruta
Updated over a week ago

On Pulsar, we believe in helping people understand conversations and public discourse, however traditional approaches to data analysis don't always account for the changing nature of conversations, or the different views and motivations of the groups of people taking part. That's why we've introduced Narratives on TRAC, a data analysis approach that acknowledges that conversations and views about a specific topic are not homogenous, albeit still connected, and that different people in that same conversation will almost always discuss the same topic differently. This teaches us to think about these conversations in terms of clearly defined narratives, instead of just an abstract collection of opinions.

How does Narrative analysis work?

Narrative analysis leverages our existing conversation clustering, popularly known on TRAC as Segments. Firstly, we analyse the dataset to extract the top 5 themes that are dominant in the search, as shown below, providing you with a succinct summary of the entire conversation, at any given time. This is keeping in line with one of the recurring themes of Pulsar 3.0 - we do all the heavy lifting for you, so that you don’t have to!

Clicking on any of the identified narratives will filter the page down by that specific narrative, giving you a more detailed breakdown of the narrative, including the scale of the conversation in that narrative, the audience behind that narrative, and how it moves across different platforms, over a given period of time.

Next, we analyse the content that is not only dominant in each narrative but is also unique to each narrative, to get a sense of the different conversations that are driving each single narrative. Additionally, you can also understand conversation dominance by looking at the most engaged content in that narrative, content with the highest media reach and content that is most impactful, i.e. has the highest visibility. Understanding content this way is a methodology unique to Pulsar, and each lens applied will always paint a different picture of how the content in each narrative differs.

Audience segmentation is a recurring theme on Pulsar, whether it’s understanding audience by communities or audience by network engagement. With Narratives on TRAC, we give you a new way of understanding the audience, which is understanding the audience of each narrative. This means we analyse the audience makeup of each narrative, to give you a succinct breakdown of their demographic characteristics and interests, as shown below.

Lastly, we identify the people who are most influential in driving each narrative. On Pulsar, influential people are defined as people who not only produce a high volume of content but also have a greater than average visibility or exposure, and exert influence around a particular subject - in this case, around a particular narrative.

This means with Narratives, you'll be able to:

1) Identify the dominant narratives in your search or campaign.

2) Measure the scale of the conversation in each narrative, and discern how each narrative spreads across different platforms.

3) Understand the main drivers of each narrative.

4) And identify the main voices behind each narratives, along with their motivations.

Narrative Analysis is available on the new TRAC in beta, and as we refine this new data analysis approach, we are keen to get your feedback on how you would use this and what additional insights you’d like to see. You can submit any feedback via the in-app chat or email product directly on

Coming Up on Narratives

  1. Narrative Summarisation: We are exploring summarisation, using tools like ChatGPT so keep your eyes peeled for this next iteration!

  2. Narrative Evolution: This will teach us to think about narratives in terms of sequences, rather than just an abstract collection of views and opinions, meaning you will be able to keep track of how public opinion today is different from yesterday or last month, and how it got there.

Did this answer your question?