We are pleased to announce that clients with an Audiense subscription can now choose how they want to segment their Communities on TRAC - either by Affinities Clustering or Interconnections Clustering, directly from Pulsar! This new update gives you greater flexibility in how you want to analyse your audiences on Pulsar, unlocking new insights that you can leverage to better understand the different audiences of any topic.
So what’s changed?
Starting today, users will be able to go into their Communities tab on TRAC and when choosing to segment an audience, they will be presented with 2 options: to segment that audience by Affinity or by Interconnectivity. Prior to this release, the default clustering method when analysing Communities on TRAC was Interconnectivity. Now, to maintain parity with what’s available in your Audiense account, and to give you more flexibility on how you want to segment your Communities on TRAC, we have introduced the option to choose your preferred clustering method! This means you can now easily switch between methods based on your needs, whether you're focused on shared interests or network-based connections. And what’s more, you can now gain deeper insight by analysing the same audience with both clustering options and comparing the results side by side -all within Pulsar!
Affinities vs Interconnections Clustering
Audiense offers two primary audience segmentation methods: Affinities and Interconnections Clustering. Below is a high-level overview of each method and typical use-case.
Affinities Clustering
Basis: Clusters individuals based on shared interests inferred from their following patterns.
Methodology: Utilises unsupervised clustering to analyze the accounts users follow, identifying patterns that indicate common interests. Users are grouped into clusters reflecting these shared affinities.
Use Cases: Ideal for understanding audience interests, informing content creation, media planning, influencer partnerships, and personalised marketing strategies.
Interconnections Clustering
Basis: Clusters individuals based on mutual relationships and how they are interconnected.
Methodology: Employs the Louvain algorithm to analyze who follows whom within the audience, grouping individuals into communities based on their direct connections.
Use Cases: Suited for discovering niche sub-communities, analysing location-specific groups, professional cohorts, and gaining a general market overview.
Key Differences
Clustering Criteria: Affinities Clustering focuses on shared interests derived from following behaviors, while Interconnections Clustering centers on direct relationships and mutual connections.
Insights Generated: Affinities Clustering provides insights into audience passions and interests, aiding in content and marketing personalisation. Interconnections Clustering reveals how audience members are connected, useful for community detection and understanding network structures.
Creating Reports
Once your Audiense subscription has been confirmed, head over to your desired TRAC search → Audience Insights → Communities page. On this page, you’ll be able to view a snapshot of the current audience makeup of your TRAC search.
On the top right corner of this page are Pulsar filters, which will give you the option to apply any filters to your current audience before segmenting it via Audiense. For example you may only want to segment the audience sharing a specific hashtag, not the entire audience of the search, or you may want to only segment the female audience in the search and not the entire audience.
Once you’re happy with the audience you want to segment, you can now proceed with creating your report as shown below. Name your report, choose your preferred segmentation method, hit the Segment button, and voila!
After a few minutes, your Report will be available on your TRAC dashboard, as shown below, with an indication of what clustering method was used to produce your Communities.
Availability
To segment your audiences on Pulsar using either Affinities Clustering or Interconnections Clustering, an Audiense subscription is needed. If you don’t have access and would like to learn more, please reach out to your Account Manager.