Comscore extended its transcript-level classification technology to five streaming and podcast platforms on July 22, 2026, giving advertisers episode-specific targeting, brand safety controls and measurement across Spotify, SiriusXM, Triton Digital, Acast and Libsyn.

Comscore announced integrations on July 22, 2026 connecting Spotify, SiriusXM, Triton Digital, Acast and Libsyn to the transcript-level intelligence built by Proximic, its programmatic targeting division. According to the announcement, issued from New York, advertisers can now activate brand safety solutions and AI-powered Predictive Audiences across streaming and podcast inventory on the five platforms, bringing audio closer to the targeting precision available in display, video and CTV.

One detail of the announcement requires clarification. Comscore's press materials render the fifth partner as "Libysn" throughout. The podcast hosting and advertising company operates as Libsyn, the trading name of Liberated Syndication Inc.

The announcement carries no financial terms, no rollout schedule beyond immediate availability, and no detail on language or market coverage. What it does describe is a shift in the unit of analysis for audio advertising: from the show to the individual episode, and from genre labels to the words actually spoken in a recording.

From genre tags to spoken words

Audio has long been bought against blunt signals. A programmatic buyer could select a comedy podcast or a news network, but rarely knew what a specific episode discussed before an ad ran against it. According to Comscore, the expanded Proximic offering moves beyond broad genre tags and demographic proxies to analyze podcast and streaming content at the transcript level, allowing campaigns to align with the actual topics and context of specific episodes rather than a show's broader themes.

Episode-level brand safety and suitability classifications sit at the center of the offering. According to Comscore, classifying individual recordings creates opportunities for more relevant placements and for brand protection grounded in what is being said in a specific episode. The distinction matters in practice. A show filed under a safe genre can devote a single episode to subject matter an advertiser wants to avoid, while blanket genre exclusions remove thousands of episodes that would have suited the brand perfectly well.

"Brands recognize that podcasts are having a cultural moment," said Jessica Trainor, Head of Partnerships, Proximic by Comscore. "By delivering transcript-level intelligence based on what is actually being said during an episode, Proximic is helping advertisers invest with confidence and reach listeners in the moments and environments that matter."

The July announcement builds on infrastructure Comscore assembled earlier in the year. Proximic's audio activation reached The Trade Desk in January 2026, delivering content-level classification across streaming inventory and 4.6 million podcasts through The Trade Desk without cookies or user identifiers. The new integrations move that classification layer from a single demand-side platform onto the supply side, embedding it directly with the companies that host, distribute and monetize the audio itself.

Five platforms with different stakes

Each of the five partners arrives at the integration from a different position in the audio supply chain, and each has spent the past two years building programmatic capacity that transcript-level signals can now plug into.

Spotify has pursued the most aggressive expansion. The company opened automated podcast buying to roughly 170 million listeners across 12 markets in July 2025 through its Ad Exchange and Ads Manager, and added podcast inventory to its direct Amazon DSP integration on July 6, 2026. No Spotify executive is quoted in Comscore's announcement, though the platform headlines the partner list.

SiriusXM brings both its media network and AdsWizz, the ad technology subsidiary that powers dynamic ad insertion across its portfolio. "Audio advertising is scaling quickly because the infrastructure around it is rapidly developing - from premium inventory and programmatic activation to audience targeting, measurement and brand suitability," said Sherene Hilal, Chief Ad Product and Technology Officer, SiriusXM and AdsWizz. "SiriusXM Media and AdsWizz were early supporters of Proximic by Comscore's audio capabilities, validating the value of these tools with advertisers and platform customers from the start. By bringing the rich signal from transcript-level intelligence to a global platform integration, we give buyers even more context, control and confidence as they activate across tens of millions of premium podcast episodes globally."

The AdsWizz connection has history. The platform integrated PodScribe's transcription-based contextual targeting in 2023, an early attempt to target podcasts by keywords and topics extracted from speech. Its own State of Audio Adtech Report later documented that contextual approaches account for 60% of all global targeting dimensions across AdsWizz platforms, a figure that helps explain why a contextual targeting vendor would find receptive partners in the SiriusXM ecosystem.

Acast, which describes itself as the world's largest pure-play podcast company, connects more than 140,000 creators with more than 4,000 advertisers globally, according to the company, and trades on Nasdaq Stockholm under the ticker ACAST. The company ran the first integrated video ad campaigns on Apple Podcasts in May 2026, and has positioned granular buying controls as a competitive differentiator. "Podcast advertising isn't just an afterthought anymore; it's a core buy," said Valerie Reimer, SVP Business Development & Strategic Partnerships at Acast. "By bringing Proximic by Comscore's transcript-level intelligence to Acast, we're giving advertisers the granular control they've been asking for. It allows brands to activate across our inventory with the absolute contextual confidence they've come to expect."

Libsyn, one of the longest-running podcast hosting companies, signed five audio and video podcasts to exclusive hosting and advertising deals in April 2026, expanding the inventory its Libsyn Ads business sells across content verticals. The Comscore integration adds a classification layer to that inventory. No Libsyn executive is quoted in the announcement.

Triton's dual role

Triton Digital occupies the most intriguing position among the five. The company operates in more than 80 countries and runs Webcast Metrics for streaming audio measurement alongside Podcast Metrics, one of the first IAB certified podcast measurement services, according to the company. It also owns a competing transcript-analysis business. Triton acquired the audio intelligence platform Sounder in March 2024, pushed Sounder.AI contextual segments into The Trade Desk on May 20, 2026, and added sentiment scoring to the product on July 2, 2026.

Enabling a rival classifier alongside its own suggests Triton sees more value in making its inventory legible to as many demand sources as possible than in protecting a proprietary signal. "Podcasting is a very high engagement media environment, and advertisers increasingly want to make it a more meaningful part of their plans," said John Rosso, President and CEO, Triton Digital. "But the medium is highly fragmented, and buyers need better ways to identify the right content and context for their message. Triton is pleased to enable Comscore's transcript-level intelligence, helping advertisers bring greater precision, relevance and confidence to their audio investments."

The numbers behind the pitch

Comscore's case rests on an engagement figure from its own research. Among 25-34 year olds, 45% tune into podcasts each week, the highest rate of any age group, according to Comscore Plan Metrix Multi-Platform data covering desktop and mobile usage among adults 18 and older, with fieldwork in April 2026. The figure comes from a company that sells targeting products into the market it is measuring, a point worth keeping in view when weighing the claim.

Independent industry data points in a similar direction, however. Podcast advertising spending climbed 32% year over year in the fourth quarter of 2025, with 1,482 brands entering the channel for the first time, according to Magellan AI benchmark data. Yet the structural gap Comscore describes is real. Consumers dedicate 31% of their media time to audio while advertisers allocate only 9% of budgets to the channel, a disparity that has persisted across multiple measurement cycles. According to Comscore, investment has lagged consumer engagement because buyers lacked tools for relevance, suitability and accountability comparable to other programmatic channels.

Whether tooling alone explains the shortfall remains an open question. Research published in December 2025 complicates the picture: the Global Podcast Advertising Compass, a study spanning 50,000 podcasts in five markets, found that existing brand safety filters were already blocking engaged listeners rather than protecting budgets. Granular classification cuts both ways. Precise episode-level signals can open inventory that genre exclusions would have discarded, or they can multiply the number of ways an episode gets flagged. The outcome depends on how advertisers configure the thresholds, and on how accurately the classifiers read conversational speech.

A crowded race to read podcast transcripts

Comscore is not alone in treating the episode transcript as the new unit of audio ad decisioning. Integral Ad Science brought episode-level pre-bid suitability controls to Spotify podcast inventory in June 2026, routing its classifications through The Trade Desk. Triton's Sounder, as noted, reached the same demand-side platform in May. The pattern across all three announcements is consistent: transcription, natural language analysis of conversational speech, and classification into topics, suitability tiers and, increasingly, sentiment.

What distinguishes the July 22 announcement is breadth on the supply side. Rather than surfacing signals inside one buying platform, Comscore has embedded its classification with five companies that collectively host, distribute or monetize a substantial share of the open podcast ecosystem, spanning Spotify's logged-in audience, SiriusXM's streaming network, Acast's independent creator marketplace, Libsyn's hosting base and Triton's broadcaster-heavy infrastructure.

Why the announcement matters

For media buyers, the practical change is that episode-level suitability and contextual segments become available where audio inventory is sold, not only where it is bought. That narrows a long-standing asymmetry between audio and other programmatic channels, where content-level signals have informed bidding decisions for years. It also standardizes the classification vocabulary across five platforms that previously exposed different levels of content detail, which reduces the operational cost of running consistent suitability frameworks across an audio plan.

For publishers and hosting platforms, transcript-level classification changes the economics of individual episodes. Inventory that genre-level blocking previously excluded can re-enter campaigns when the actual content proves suitable, while episodes touching sensitive subjects face more precise exclusion. For the measurement business itself, the announcement extends a pattern visible across Comscore's recent moves in audio, political CTV and television tracking: classification and prediction products layered onto the company's measurement footprint, sold through integrations rather than as standalone tools.

The unanswered questions are commercial. The announcement does not state what advertisers pay for the segments, which languages the transcript analysis covers, or whether any platform receives preferential access. Nor does it address how the classifications will be audited. Those details will determine whether transcript-level buying becomes the default for audio, as content-level signals did for display, or remains a premium option layered onto a channel still fighting for its share of the budget.

Timeline

Summary

Who: Comscore and its programmatic targeting division Proximic by Comscore, in partnership with Spotify, SiriusXM and its AdsWizz subsidiary, Triton Digital, Acast and Libsyn. Executives quoted include Jessica Trainor of Proximic by Comscore, Sherene Hilal of SiriusXM and AdsWizz, Valerie Reimer of Acast and John Rosso of Triton Digital.

What: Integrations that make Proximic's transcript-level intelligence available across the five platforms, enabling advertisers to activate episode-level brand safety classifications, suitability controls and AI-powered Predictive Audiences programmatically across streaming audio and podcast inventory, replacing show-level genre tags with analysis of what is spoken in individual episodes.

When: Announced on July 22, 2026, with the capabilities described as available immediately. The move follows Proximic's January 2026 audio activation inside The Trade Desk and a series of competing episode-level classification launches from Triton's Sounder and Integral Ad Science during the first half of 2026.

Where: Announced from New York. The integrations span the five platforms' global streaming and podcast inventory, with SiriusXM citing tens of millions of premium podcast episodes; the announcement does not specify language or market coverage.

Why: According to Comscore, podcast investment has lagged listener engagement, which reaches 45% weekly among 25-34 year olds per the company's April 2026 Plan Metrix data, because advertisers lacked relevance, suitability and accountability tools comparable to display, video and CTV. Embedding transcript-level classification with the platforms that sell audio inventory is positioned as the mechanism to close that gap.