Audio Insights Dashboard

Overview

The Audio Insights Dashboard gives you a new level of insight into your content with our brand suitability and contextual targeting solution. With the dashboard, you can:

  • Assess the suitability of podcasts before you put them on a media plan
  • Discover podcasts you may not have considered by reviewing IAB category and suitability risk levels.
  • Use Sounder's Brand Safety Badge to confidently offer your content to customers concerned about brand safety and suitability.

Not all podcast content is created equal, and Sounder’s AI solutions solve for precision and trust challenges the industry faces when buying podcast inventory at scale:

  • Avoid highly objectionable content that can damage brand reputation
  • Identify content that positively aligns with a brand’s audience and values
  • Deliver campaigns to safer, more suitable inventory that drives meaningful campaign results

Sounder’s Signal

When Sounder processes a podcast, we provide insights in three main areas:

  • Contextual Targeting: We create precision targeting and scale using contextual targeting to go beyond a show’s assigned genre. For example, a self-described Sports podcast may also discuss Entertainment or Business in many of its episodes.
  • Brand Safety: This is now table stakes for advertisers and the podcast industry, so on principle, we exclude any inventory that is classified as "Floor" and therefore unsafe for any advertising.
  • Brand Suitability: Brands have varying risk tolerance levels for their brand and on a per-campaign basis. We enable buyers to set their preferred suitability risk levels so the brand’s message only runs across inventory that is suitable to their preferred risk levels.

We built natural language processing models using an ML/AI approach to understand the spoken word dynamically. We don’t look at keyword density; we analyze the actual context of the content to understand the speaker's intended meaning.

Using our proprietary Automatic Speech Recognition (ASR) model, we transcribe the content of a podcast and utilize our GARM/IAB brand safety & suitability models and IAB Content Taxonomy Contextual model to surface the resulting content classifications through Audio Insights.

We’re a third-party verification tool for brand safety and suitability.


Who is the Dashboard for?

Several type of customers will find dashboard useful:

  • Publishers and networks who would like brand suitability and contextual insights into their content
  • Media Agencies looking for suitable inventory for clients
  • Partners seeking to enrich their content with Sounder's signal

How to get access

If you’re ready to see how Sounder’s solutions can transform the power of your inventory, contact info@sounder.fm to get started!


Key features

The Audio Insights dashboard provides deep brand suitability and contextual insights into your podcasts. The key features include:

  • Show-level brand safety and suitability scores and contextual topic analysis
    • See a top-down view of your catalog
  • Episode-level brand safety and suitability scores and contextual topic analysis which not only provides the details behind the show-level scores, but also offers a way to find new inventory related to content categories
  • Episode details: Segment-level brand safety and suitability scores and contextual topics so you can see the context behind the scores; Exportable transcripts; audio playback; heat map navigation directly to segments related to brand suitability categories;
  • Click to instantly filter your content by risk level, contextual category, and named entities across podcasts, episodes, and segments.
  • Segment Builder for advanced filtering
    • We've made some changes to how filtering works to create a unified experience. Stand-alone GARM category filters at the top of the page have been removed in favor of a single Filters button that opens the Segment Builder window.
    • To get started, click on the Filters button in the top-left corner of the dashboard.

    • The system will add a default filter for brand safety and suitability (GARM) that includes all inventory. From there, you can customize GARM filters and add in filters for IAB contextual topics and named entities (people, places, things).


    • Dimensions. When you start to build a filter, you'll need to select a Dimension. Dimensions are the types of filter you can apply to your content. Using "All" and "Any" logic, you can build complex expressions across the different types to define very specific segments of your inventory. We will continue to add dimensions to improve the sophistication of your filters. When you add a filter, you can select from three options:
      • GARM - this lets you filter by brand safety and suitability scores across 11 GARM categories. You can apply the same filter to all GARM categories, or customize it by category. Learn more about the GARM framework here.
      • IAB - this is a list of 40 Tier 1 contextual topics according to the IAB Content Taxonomy 3.0
      • Named Entities - these are the people, places, and things in the content that have been identified by Sounder's technology.


    • Expressions and "All" or "Any" logic. Your filter can include or or more expressions depending on how wide or narrow a target you are trying to reach. The use of "All" or "Any" logic within and across expressions gives you a powerful tool to create inventory segments tailored to your needs.
      • “All” corresponds to an “And” expression, so all the criteria must be true in order for an episode to be included in the filter.
      • “Any” corresponds to an “Or” expression, so any of the criteria may be true for an episode to be included in the filter.
      • In the example below, the red selection controls the relationship across filters. The goal is to find shows and episodes where:
        • GARM is Unrelated or Low for Crime, Obscenity, and Adult categories, but Medium or lower for all other categories AND
        • Named Entity is "NBA" AND
        • Named Entity is "Denver Nuggets" OR "Nikola Jokic" OR "Jamal Murray" OR "Michael Malone"

          Example of "All" and "Any" logic
        • For episodes to be included in the filter, the GARM scores for all episodes must be in the defined ranges AND mention "NBA" AND mention at least one of the following: "Nuggets", "Jokic", "Murray" and "Malone".
        • Within the Named Entity filter, the “Any” selection will enable the filter to capture any episode where at least one of those named entities is present.
        • The Duplicate and Delete icons on the right of each filter allow you to quickly build more customized expressions. In the example above, using the Duplicate button makes it easy to keep most of the selections the same, and only requiring the GARM category to be changed.
        • Click the Apply button to return to Dashboard with the filter in place.


    • Managing Filters
      • There are several actions you can take with filters:


      • Save a filter. After you Apply your filter, the system gives it a default name like "QueryFilter1". You can name the filter and save it by clicking on the "Save View" button.
      • Find and re-apply a filter. Use the All Views button to re-apply saved views (or filters)

      • Edit a filter - Reset or Save Edits
    • Export CSV

      Export your data to CSV
      • Assuming you have the appropriate permissions, you can export your GARM brand suitability data and your top five contextual topics per show
        • Both exports will apply any filter that is active when you start the export. Remove any filters to export data for all shows.
        • The contextual export includes the top five (5) ranked IAB contextual topics for each show that exceed a 70% confidence score. Some show may have fewer than five topics, or none, depending on the model's confidence in the results.

  • Megaphone Integration
    • For customers that use Megaphone, we offer an integration that makes it easy to use your Sounder data directly in Megaphone Orders with Advertising Tags.
    • See this article for more information, or contact us.
Did this answer your question? Thanks for the feedback There was a problem submitting your feedback. Please try again later.

Still need help? Contact Us Contact Us