Conversations: tracking what is said about brands in spoken media
Conversations adds spoken media to Brand Signal: brand mentions in podcasts, on YouTube and TikTok and in broadcast, measured next to survey data and built with All Ears.
Conversations adds spoken media to Brand Signal. Alongside survey data on what people think of a brand, it tracks where the brand is mentioned in podcasts, on YouTube and TikTok, and on radio and TV, using data from All Ears.
Brand tracking has long measured what people think of a brand, with survey measures such as awareness, consideration, brand equity and positioning. Surveys do not show where a brand comes up unprompted in podcasts, on video platforms and in broadcast.
Brand teams have mostly been unable to follow those mentions, or have used separate tools that cover written text.
Three kinds of brand data in Brand Signal
Perception is what people think, measured in survey-based brand tracking, competitive mapping and positioning.
Conversation is what people say, measured as mentions of brands across podcasts, creators, social video and broadcast.
Visibility is what AI says, measured as how often AI assistants mention a brand and how they describe it.
All three are in the same platform, cover the same brands and markets, and can be queried in MixerMind. Survey data remains the foundation, and the other two show whether a change in perception also appears outside the survey.
What Conversations measures in spoken media
For every brand mention, Conversations records four things:
Mentions, the number of mentions and the brand's share of all mentions in the category, compared with competitors.
Reach, the estimated audience the mentions reached and the brand's share of total reach.
PR value, an estimate of the equivalent media value of the mentions, based on reach, sentiment and channel.
Sentiment, whether mentions are positive, neutral or negative, summarised in a Net Sentiment Score.
Overview of total reach and PR value for a brand.
Behind every number is the mention feed itself, with the quote, the channel, the source and the sentiment of each mention.
How brand mentions are checked
Spoken media is noisy, so each mention passes several checks, done with AI models, before it is counted:
Market attribution. A Swedish-speaking podcast may come from Sweden or Finland, and a Spanish-speaking creator may address several markets. Each mention is assigned to the most likely market, also where language alone does not tell.
Topic relevance. Each mention is scored for relevance to the category, so that apples the fruit are not counted as Apple the brand. Off-topic mentions are kept for audit but left out of the reported numbers.
Brand validation. Brand names are checked in context, together with the topic score, to confirm that a mention refers to the right brand.
Sponsored or organic. Mentions inside ad reads, sponsored segments and paid collaborations are flagged, so paid and earned reach can be reported separately.
Spoken media data in MixerMind
MixerMind queries the conversation data directly, alongside the survey metrics. A question such as what drove a rise in reach last month, or how often a brand is mentioned compared with competitors and in which channels, is answered from the same platform.
The partnership with All Ears
Conversations is built on a partnership with All Ears, a Swedish company that indexes and structures spoken media at scale and processes around 50,000 hours of audio a day. All Ears provides the transcribed mentions through an API.
Odyssey adds market attribution, validation and the analysis inside Brand Signal, and connects the mentions to its segmentation and positioning models.
”Traditional brand tracking has been reactive for far too long, telling you where a brand has been instead of where it’s heading. By integrating All Ears’ real-time spoken media data directly into Brand Signal, we are killing that lag. Brands can now stop guessing and start watching narratives form as they happen across podcasts, streaming, and social media. This completely shifts the playing field for data-driven brand strategy.”
— Tobias Sjöqvist, CEO, Odyssey AB
"Together with Odyssey, brand tracking moves from reactive to proactive. With structured tracking data and live spoken media in the same view, brand leaders no longer just measure where they've been, they see where they're heading, and can shape narratives as they form rather than respond to them after the fact."
All Ears provides spoken media intelligence across podcasts, social media, broadcast and web TV in multiple languages. Its platform delivers structured, real-time spoken media data for AI systems, APIs and intelligence workflows, and its SaaS platform is used by brands including Walmart, KFC and Hertz.
