Introducing the Next Generation of Audience Analysis in Brandwatch

Now, how about some trickier panels?

• Tesla owners
• Tesla brand advocates 
• Tesla brand detractors

To create these panels, we’re going to need a different approach. A bio search won’t cut it. We can start by creating a query to collect all mentions of Tesla across the web.

Once we have all this data, we can refine the search to tag all conversations where people specifically talk about being a Tesla owner or driver. We have all the boolean operators in Brandwatch at our disposal to create a tag looking for conversations like “I AND (drive OR bought OR own) NEAR (tesla OR “cyber truck” OR “model 3”).”

Now, if we filter our dashboard by this tag, we have a set of conversations by authors on many different platforms about being a Tesla owner. And with the click of a button, we can turn all the authors in this dashboard into a new Social Panel.

But advocates and detractors are going to be even harder. We could filter our data by sentiment but this is likely problematic. Many people might mention Tesla in a post that overall is negative or positive, but the sentiment is not aimed at the Tesla brand. Some might just be big fans of ‘Tesla CEO Elon Musk’. Some might be disappointed owners of Tesla stock when it’s taken a dip.

Fortunately, we have all of Brandwatch’s AI-powered analytics tools at our disposal to zero in on the right conversations to build the best, most relevant panels. In this case, we’ll use Custom Classifiers, our machine learning tool that trains a custom classification model by dragging and dropping mentions into different categories.

This lets us sort conversations into buckets for irrelevant or neutral posts, posts that are positive or negative but not in relation to Tesla, and finally, the ones we want: positive or negative posts about the brand itself.

Again, we can filter our dashboard by these two final categories to create our last two panels.

Breaking down data by panels 

Now, we can go back to our original brand query and break the conversations down by the panels we identified.

This allows us to build a far more nuanced ongoing understanding of how different audiences talk about our brand. We can set up a volume over time chart broken down by these panels to monitor when certain audiences discuss the brand.

Iris, Brandwatch’s AI assistant, is particularly useful here as it will detect and explain peaks for each of these social panels. For example, Iris has identified the big spike as brand advocates talking about Tesla and it’s driven by reposts of this post that Elon Musk shared in August.

Iris is even more helpful in identifying peaks like this where the human eye would miss it. For example, peak A detected here is within the investor conversation, which is a much smaller volume because it’s a smaller panel of people. We can see it better once we remove a few panels.

If we ask Iris to summarize this peak, we can see that it has identified a flurry of activity within this community around Tesla’s Q2 earnings report.

We can use this monitoring broken down by panel to identify moments of opportunity and risk for the brand and cut through the noise to zoom in on the trends within the audiences we care most about.

As well as monitoring over time, we can dig into and compare how each of these panels talk about the brand. For example, we can categorize conversations using keyword rules to break down by key topics we want to learn more about.

In Consumer Research, you can chart social panels against these categories or any other segmentation of data with complete flexibility to build reports that reveal insights and tell compelling data stories.

Here, for example, we can see that environmentalists are most likely to be talking about Tesla in the context of sustainability, investors talk most about price, and brand advocates and owners talk about the aesthetic and design of Tesla vehicles.

For any of these topics, we can drill down and compare how each audience is talking and better understand how messaging is pulling through to these key audiences and to learn how to better engage them around key topics.

Researching panels

We can also dive deeper into the panels we have built and understand everything they talk about, even when they are not talking about Tesla.