Alternative data for hedge funds and asset managers
Podcast and newsletter data for financial institutions
Every mention of the companies you cover across podcasts, Substack and Bluesky, keyed to ticker and CIK, with the passage, the speaker and the time it became available.
featured in
Hedge funds, asset managers and research teams use alternative data to see what the filings and price feeds don't show. Some of the most useful commentary on a company is spoken: a CEO on a business podcast three weeks before earnings, an analyst on a sector show, a former employee on a long-form interview. Newsletter writers on Substack and posters on Bluesky add their own takes within hours.
That material is audio and long-form text, so it's hard to search, join to a security master or backtest. Subalytics transcribes new episodes from 250K+ tracked podcasts every day, identifies who is speaking, and matches every mention of a monitored public company across podcasts, Substack and Bluesky. You get rows you can query, keyed to ticker and CIK.
What you get
Three datasets that join on stable IDs
Company mentions
/api/v1/entity-mentions/
One row per mention of a monitored company in a podcast, Substack post or Bluesky post: the snippet, how it matched (cashtag, name, alias, product or ticker), a confidence score, a link to the source and, for podcasts, the start and end time in the audio. Daily or weekly counts per company and trending rankings by volume or change.
Podcast transcripts
/api/v1/podcast-transcripts/
Full timestamped transcripts of new episodes, split into segments, with each speaker's name, role, title and company where they can be identified.
Executive interviews
/api/v1/podcast-transcripts/
Because speakers carry a title and a company, you can find podcast appearances by a company's CEO, CFO or other executives, then read what they said around a specific moment.
Creator and audience data
/api/v1/subscribers-historical/
Subscriber and follower history for the newsletters, accounts and shows behind each mention, joined on sub_id, so you can weight a signal by reach and growth.
Beyond the earnings call
Management talks between calls, and everyone else talks every day
Earnings-call transcripts capture four scheduled conversations a year. Executives also appear on podcasts to discuss strategy, products and hiring, often in more candid terms. Between those appearances, hosts, guests and newsletter writers discuss the same companies: suppliers, customers, competitors, former employees and analysts.
Subalytics covers both sides. You can follow what a company's own executives say and measure how often, where and by whom the company is discussed. Everything is delivered through an API, so the data goes straight into research notebooks, models and internal tools.
How research teams use it
Track narrative shifts
Read the passages behind a change in how a company is discussed, with the speaker and source for each, so an analyst can read exactly what was said.
Catch events early
Mention counts are available daily or weekly and split by source. A spike in podcast mentions of a supplier or product shows up as a number you can alert on.
Monitor a coverage list
Pull mentions for your tickers on a schedule. Bluesky mentions arrive within minutes and podcast mentions within hours of the daily 05:00 UTC transcription run.
Feed models and agents
JSON over a documented REST API (OpenAPI 3.1), with stable IDs and timestamps, so the data can be used as features or as context for internal AI tools.
Built for due diligence
The questions a data team asks before a trial
| Topic | What you get | Details |
|---|---|---|
Point in time | published_utc, created_at, transcribed_at, updated_at | When content was published, when Subalytics first found it, when a transcript became available and when a record last changed. Backtest on availability time, not publish time. |
Join keys | symbol, cik, sub_id, mention_id | Ticker and SEC CIK for companies; stable IDs for creators, podcasts, episodes and mentions. |
Matching | match_type, confidence | Cashtag, company name, alias, product or ticker. Confidence from 0.5 to 1.0; mention counts use 0.7 and above. |
Corrections | updated_at, periodic re-pull | Past content is re-checked, so a mention can be added, re-scored or removed. The docs explain how to stay in sync. |
Sources | content_url | Every record links to the public podcast episode or post it came from. |
Coverage and history
Subalytics tracks 250K+ podcasts. It's a curated list of news, business, sports and technology shows, chosen because that's where decision-makers talk, and more shows can be added on request. Company mentions also cover Substack and Bluesky, and the monitored list includes 20,000+ public companies with ticker, CIK, exchange, sector and industry.
The standard API serves episodes from when tracking began for each podcast. Historical data can be provided on request.
Public sources
Subalytics collects public, on-the-record content: published podcast episodes, public Substack posts and public Bluesky posts. Each record links back to its source, and a data dictionary is available for every endpoint.
FAQ
Common questions
No. Subalytics collects published podcast episodes and public Substack and Bluesky posts. Every record includes a link to the original episode or post, so your compliance team can check any data point against its public source.
Use the availability timestamps. Transcripts carry transcribed_at and mentions carry created_at, which record when the data existed in Subalytics. Filtering on those instead of the publish date keeps a backtest to what you could have known at the time.
The standard API starts from when each podcast was first tracked. Historical data can be provided on request; tell us your coverage list and the period you need, and we'll confirm what's available.
The monitored list is available through the entity-companies endpoint, with ticker, CIK, exchange, sector and industry. Coverage is broken down by sector and exchange on the coverage page.
Each mention records how it matched (cashtag, company name, alias, product or ticker) and a confidence score from 0.5 to 1.0. Daily and weekly counts include only mentions at 0.7 or above. Past content is re-checked as matching improves.
Yes. Send us your coverage list and we'll set up access to the endpoints you need so your team can evaluate the data against its own research process.
Contact us
Have a question or a custom use case?
Tell us what you're trying to monitor or search across, and we'll get back to you within a business day.
Test it on your coverage list
Tell us which companies and sources you follow. We'll enable the matching endpoints and walk your data team through the fields.