Podscribe Turns To Audio Signals To Help Solve YouTube Performance Puzzle.

As more podcasters distribute on YouTube, the podcast attribution challenge has been growing for Podscribe as it works to determine the effectiveness of advertising on series that may not even have an audio or RSS component. To help fill the gap, the company has introduced a new YouTube-only attribution model that it says offers brands a way to measure campaign performance for YouTube-only influencer placements.

Podscribe CEO Pete Birsinger says YouTube remains the most difficult video platform for the company to measure because it can’t obtain the IP addresses needed for direct attribution.

“By and large, YouTube we can do no form of true attribution,” Birsinger said during a webinar Wednesday. “We cannot get the IPs right now.”

Instead, the new data offers modeled estimates rather than direct user-level attribution. Podscribe extrapolates expected YouTube outcomes using comparable performance observed from a campaign’s attributable audio impressions, combined with signals specific to the video and ad.

“We can take the podcast response rate, then apply that to the YouTube views, and then weight it a bit up or down depending on the YouTube channel’s subscriber location, the comment-to-view ratio, the location of the ad in the YouTube video, and a number of other factors,” Birsinger says. “We can model for it, but it’s still modeling.”

The approach doesn’t allow Podscribe to determine which individual viewers took action after seeing a YouTube ad. But the company says the probabilistic model gives advertisers a practical way to understand how YouTube placements may be performing.

It also means an advertiser needs attributable audio campaign data that Podscribe can use as a baseline. Without comparable audio performance, the company has no foundation on which to model the YouTube-only placement.

When it has both, Podscribe can estimate a brand’s incremental response based on where an ad appears in a video and the engagement rate for the individual video, measured in part by how many comments it receives. The model also incorporates response signals such as promo codes and vanity URLs.

Podscribe says those elements allow it to identify meaningful performance differences among YouTube videos. Senior Director of Partnerships Camden Weber says the company has seen shows generate plenty of YouTube views but little modeled response when viewers are not actively engaging with the content.

“What we observed was that there were one to three comments per YouTube episode,” Weber says. “There were many views on it, but really not a very engaged audience, and the performance reflected that.”

The model arrives as video podcast distribution spreads across YouTube, Spotify, Apple Podcasts and Rumble, creating what Birsinger calls an increasingly fragmented measurement landscape. He says YouTube stands out as the platform where Podscribe’s attribution capabilities remain “the most limited” because it can pull view counts but not the user-level data available through some other distribution systems.

The company says in the announcement that YouTube-only modeled attribution will be automatically available for eligible pure-YouTube campaigns at no additional charge through Oct. 31.