All Collections
Use Case Guides
TRAC Use Case: Exploring a Category Search
TRAC Use Case: Exploring a Category Search

This article provides a step-by-step guide on how to create and explore a category search within Pulsar TRAC.

Madeleine Day avatar
Written by Madeleine Day
Updated over a week ago

Setting up a category search can be really exciting as it opens up exploratory insights such as trend analysis, consumer behaviours, true share of voice and loads more. It allows you to look beyond a specific brand search or a niche subject matter to find rich insights and learnings from the wider conversation that you might have otherwise missed.

An example of a category search would be, rather than tracking a particular gin brand's online mentions, you looked at how gin or spirits were talked about. Or if you were focused on discussion around wind farms, you'd go wider and look at the renewable energy conversation for example.

For this article, we will be using the example of exploring the gin category.

Gin Tonic Weekend GIF by HENDRICK'S GIN

Learning Outcomes:

  • Setting up a category search and how to refine this

  • Recommended charts and visuals to showcase data

  • Using the filter to pull forward insights

  • Inspiring the use of custom charts to answer key questions

Step 1: What are we searching?

For a category search you will first need to create a topics search within Pulsar TRAC.

The trick to a successful category search is to keep your keyword selection simple. What are the obvious words or hashtags someone will use when referencing the category? Do a little desk research to identify most commonly used hashtags. For the gin category of conversation, it is simply:





The first thing you'll need to to when setting up your search is select data sources. For the best and most exploratory category searches you'd pick as many data sources as possible to uncover if the category is talked about in places you didn't expect. However, depending on how broad your category topic is, it could consume A LOT of data!

🚨 Important: It's vital to monitor your data usage in order to ensure you are staying within your team's monthly allowance. The following article provides a tutorial on how to manage your Data Usage.

Therefore, keep this in mind and make sure to carefully select your data sources. For example, Gin is most likely to be mentioned by consumers across social channels like X, Facebook, Instagram and Pinterest. Commentary on certain brands and specialist gins might be better found on Forums, blogs and reviews. Also, if its important to include how the media is contributing to the online conversation then include news sources too. Think about the different angles of the conversation and what is most important for you to capture at this step.

More tips on refining your search and data volumes can be found below in step 3!

Step 2: Entering Our Keywords

Using either the wizard or boolean search set up option, input the keywords that you want to track. For the gin category, we'd track the simple keywords and popular hashtags as mentioned above. When entering them into a wizard set up, they'd look like this in the platform:

For a full walk through of setting up a wizard search, click here:

Or if inputting these keywords into boolean search set up, it would look like this:

Gin OR #gin OR #ginoclock OR #ginlovers

For a full walk through of setting up a boolean search, click here:

Step 3: Refine your Query with Operators

Category searches tend to have much larger volumes of conversation due to their broad nature. We don’t want you to be wasting your data on irrelevant content, so here are a few tips and tricks on how to refine the search.

  1. If using boolean to set up your search, using the location and language operator to narrow down the scope ie. LOCATION “GB” AND LANG “en”. If you have used the wizard, you can select these options in the 'target' step of your set up.

  2. Excluding keywords within your set up is an easy way to rid of irrelevant conversations eg. competition, win, giveaway, voucher, promo code, ad, voucher. If using boolean use the AND NOT operator.

  3. Excluding retweets, replies and quotes so you can only see original content is a good way to reduce volume ie. use this operator in your boolean AND NOT RT

  4. Using the X sample operator is a great way to capture the breath of the conversation but reduce the overall volume, as it will only pull in a random percentage of real time data collection ie. use the SAMPLE 20 operator for boolean or using the sliding scale on the 'target' step of your wizard set up.

Combining some of the above, your boolean expression would look like this:

(Gin OR #gin OR #ginoclock OR #ginlovers) AND NOT "snoop dog" AND LOCATION "US" AND LANG "EN" AND SAMPLE 20

💡 Top Tip: Use the Summary section within search setup to see a sample of the conversations you’ll be pulling into the platform. This will guide you on what to put in the exclusions section or if you need to reduce the volume of results in line with your data allowance.

Step 4: Analysing and Extracting Insights

Once you have loaded in your data, it's time to analyse and extract insights. Take advantage of TRAC's analysis modules to do some of the work for you! Utilising the topic, entities or image analysis, in conjunction with the date range filter can be used to uncover seasonal patterns, growing trends, and what are the most popular content themes driving conversation. Narratives is also a must use feature, helping you to understand that its not all one big similar conversation, but there are different conversations happening within your category search. Find out the details behind each conversation segment to help your understanding of the category.

Which charts are most helpful?

Now that your data is pulled into the platform, let's take a look at some of the best visuals to pull insights.


The Snapshot is a great way to start your analysis, as you’ll be able to extract instant insights. Using the Key Performance Metrics you can see the volume of conversation within a certain time-period and what’s driving spikes in conversation from the 'Content Over Time' graph. Don't forget to use the date range filter to focus on a wider time frame which can be helpful to analyse seasonal trends or simply see if conversation is increasing or decreasing within the category.

💡 Top Tip: Everything within Pulsar is interactive so make sure you’re clicking through on the graphs to find the next level of detail. For example on the 'Content Over Time" graph, you can hover over and click through the bubbles, which represent the top posts within your dataset, in order to see the individual posts which are driving the conversation each day.

'Trending in this search' will provide some thought provoking insights as you’ll see what are the most mentioned topics and who are the most spoken about people and organisations within this conversation, which is driven by our entities analysis. Are there any particular brands or notable names linked to the category? For example, in the gin category it is not uncommon to see Ryan Reynolds' name associated due to his link to Aviation Gin.

Topics, Entities, Image Analysis

Further dig in and utilise the topics, entities and image analysis to help identify key themes, which are found within the 'content insights' tab. The treemap visuals are great for understanding the different conversations taking place on separate channels. Does the conversation change on X vs Facebook - why might this be?

💡 Top Tip: Use the ‘Edit Data’ button to remove the generic terms from these visuals and highlight only the niche topics within the conversation.

Sometimes, by removing the elements of the conversation you know, helps you to focus on the lesser known themes and identify those magic 'unknowns'!


The Narratives tab enables you to identify that not everyone is talking about the topic in the same way. It helps you to piece together that different areas of the conversation are driven by certain audiences and individuals, on specific channels. It also helps you to understand if the narrative was a one off spike or a constant and growing conversation. For example, In the below screenshot you can see that one narrative (the green one) is focused around gin cocktail recipes specifically.


Due to the broad nature of category searches, you will need to utilise the filter to pull forward relevant topics of discussion. For example, you might have a hypothesis that certain gin flavours are beginning to become more popular. Therefore you'd be best to create a keyword filter using boolean operators like this to isolate the conversation around certain flavours:

pink OR lemon OR orange OR pear OR rhubarb

Once you’ve built out a filter, make sure to save it so you can easily access it again within 'Saved Filters'.

For more information and inspiration on filtering, click here:

What type of reporting output is best?


Creating your own custom dashboards within the platform will be the best way to collate your findings and showcase any hypothesis you might have tested against your category search.

To create a new dashboard go to Reports → Create Dashboards → Create from Scratch → Start Now.

Once you have a blank dashboard, you just need to add charts that are most helpful to visualise your findings. You can add 'standard charts' which are produced by Pulsar's visuals or you can create your own 'custom charts' to display the insights you wish to highlight. Click here for further guidance on creating custom dashboards:

Custom charts are the best way to answer questions you want to query against the data set. We'd recommend noting down the questions you are trying to answer from your category search to steer your use of filter and custom charts. For example, here are a set of questions you could apply to the gin category search:

Where are people drinking gin?

What are the moments associated to drinking gin?

What is the most popular gin cocktail?

Which gin brand have the biggest share of voice?

What is the most popular garnish/mixer for gin?

Do people reference gin as a 'healthy' drink?

What flavours of gin are most popular?

Are pink gin mentions in decline?

What audiences talk about gin the most?

When you create a custom chart, you create the filters needed to represent the answers to your questions. For example, if you wanted to understand where people are mentioning drinking Gin, you add in keyword filters to represent the different possible venues ie. garden, bar/pub, club, hotel, home, party etc.

Here are some examples, relating to the questions proposed above.

Hopefully this can inspire your use of custom charts in your category searches.

Step 5: Take Action and Explore

Armed with the insights gained from TRAC, you can now take actionable steps. Utilise the data found in a category search to explore growing themes, behaviours and query the data set to find answers to your burning questions and hypotheses. Use these insights to inform campaign planning, product development, communication strategy and so much more...

We hope you enjoyed reading this article! 📚

If you have any questions or would like to learn more, please don't hesitate to reach out to our support team via live chat. 🚀

Did this answer your question?