Founder and CEO of Podscribe - the platform that turns "we think the ad ran" into a number you can stand behind.
The Pitch
An advertiser pays to have a podcast host read their script. The episode airs. Then the awkward question: did the host actually read it - and did anyone buy anything? Peter Birsinger built a company whose entire job is to answer that, episode by episode.
Now
Podscribe transcribes every episode it tracks, labels the ads inside, and snapshots those episodes multiple times to catch exactly which spots aired and when. A pixel ties the listen to what happens next on an advertiser's website. AI does the heavy crawling; humans do the checking. The output is plain: the ad ran, here is the ROI.
It is an unglamorous corner of media - the back office of an industry that spent years selling on vibes and download counts. Birsinger leans into it. Podscribe positions itself as a third-party, IAB-aligned measurement layer for podcast and, increasingly, YouTube advertising. The promise is the same one digital marketers take for granted everywhere else: full-funnel attribution, conversion lift, and a dashboard instead of a guess.
The company started in the New York podcast scene and now operates out of Austin, Texas, with a team that has grown well past the handful Birsinger started with.
"We transcribe every episode, label ads, and snapshot episodes multiple times to track ad placements."
By the numbers
The Method
The Road Here
The Turn
The neat thing about the arc is the flip. Podible was the fun side of the business - a consumer app that learned what you liked and recommended new shows. Podscribe is the opposite end: the unglamorous infrastructure that tells advertisers whether their money did anything.
It is a very engineer move. Birsinger spent his Berkeley days on distributed computing, then years writing production code at Meraki and Tapad. Crawling thousands of podcast feeds, snapshotting them on a schedule, and reconciling what actually aired is a distributed-systems problem wearing an advertising costume.
The consumer app teaches you what people listen to. The measurement company teaches you what that listening is worth. He chose the second one.
The Findings
Measurement only matters if it tells you something you would not have guessed. One of Podscribe's clearer findings: as the ad load in a show climbs, the performance of each ad tends to fall. Stuff more spots into an episode and each one works a little less hard.
Birsinger has also tracked the long argument between baked-in ads (read once, fixed in the file) and dynamically inserted ads (swapped in on the fly). For years the conventional wisdom was blunt. Then the platform math changed.
"As ad load goes up... performance goes down. There was a definite relationship."
In His Words
We transcribe every episode, label ads, and snapshot episodes multiple times to track ad placements.
As ad load goes up... performance goes down. There was a definite relationship.
DAI used to be terrible - but with the iOS 17 update it's come a lot closer, and in some cases has even flipped.
Margins & Quirks
He collected an engineering B.S. and an applied math B.A. from Berkeley at the same time.
Both companies he founded start with "Pod" - Podible, then Podscribe. A man knows his lane.
Podscribe's contact lives on a .ai domain - a quiet nod to the machine-plus-human review at the core.
Born of the New York podcast world, the company now runs out of Austin, Texas.
His Berkeley research was parallel and distributed computing - the same mindset he points at podcast feeds.
From Meraki to Tapad to two startups, he came up writing the code before running the company.
Watch & Listen
Birsinger lays out the mechanics of podcast attribution alongside Magellan AI's Cameron Hendrix on the Podcast Perspectives series.
The Rolodex
Pass It On