Blog · 2026-09-16 · Matterfact Research

What a million podcast transcripts reveal

What a million podcast transcripts reveal about where executives appear, how company narratives differ, and where researchers can find the conversations that matter.

A company can be everywhere in the conversation and rarely have its executives in it. Apple is discussed in 32,577 episodes in this study. Its current executives appear in 239. Goldman Sachs is discussed in far fewer episodes, yet has more than twice as many executive appearances.

That distinction is easy to miss when you follow a handful of shows. Across a million podcast transcripts, it becomes a question you can investigate: how much are we hearing about a company, and where can we hear from its management?

We explored a million podcast transcripts to see what this kind of research makes possible: finding executive appearances, comparing company narratives, and tracing claims about industries, macroeconomics, and commodities.

The findings below describe this annotated sample, computed September 15, 2026. Aggregate comparisons use its full history; monthly charts end in August 2026, the last complete month.

01 — Access: management is part of the conversation

From January 2025 through August 2026, the sample averages 588 episodes featuring a current company officer per month, representing 323 companies per month.

The executive's role is checked against the episode date. Someone who joined a company's board or management team later is not counted as its current officer in an earlier episode.

The annual view starts in 2021 to show the period with broader coverage. Earlier episodes remain included in the article's all-time aggregates. The rise in captured episodes reflects changing transcript availability; it does not measure growth of the podcast market.

For a researcher, this creates a useful starting point between formal company updates. Search for appearances since your last review, then read what management discussed and compare it with the questions you still need answered. These are episode counts; they do not establish how many unique interviews or new disclosures occurred.

02 — Presence: attention and executive visibility are different

Alphabet, Apple, and Meta lead the discussion ranking. Strategy, Microsoft, and Alphabet lead the current-officer appearance ranking. Following the most discussed companies would give you a different listening list from following the most visible management teams.

Apple has about 0.7 executive appearances for every 100 discussion episodes; Goldman Sachs has about 8.1. This compares two counts. The supplied tables do not establish that every executive appearance falls within the set of episodes discussing that company, so it should not be read as the exact percentage of discussions with management present.

For company research, separate those two searches. One finds how a business is being discussed; the other finds episodes featuring its officers. For communications teams, the same comparison provides a starting point for understanding where their executives appear alongside broader attention to the company.

03 — Stance: context changes the picture

The balance of bullish and bearish company claims differs across episode groups. In the current-insider group, 51.2% of claims are bullish and 6.6% bearish. In the journalist or commentator group, 25.3% are bullish and 36.5% bearish. Neutral and mixed claims make up the remainder.

There is an essential distinction here: the annotation assigns company claims to the strongest participant tier present on the episode. It does not establish that the person in that role said each claim. These are episode-group comparisons, not quotations or individual speaker scorecards. The supplied tier table covers a subset of company claims.

“Specific” means the claim is quantified or contains a named metric, period, or source. A specific claim can still be bullish, bearish, neutral, or mixed, so specificity is measured separately from stance.

The practical next step is to compare the substance. What evidence appears in episodes featuring insiders? Which assumptions do outside commentators challenge? Where do they discuss different measures of the same business? The percentages help you choose where to look; the source material is what lets you assess the argument.

04 — Sectors: the useful conversations go far beyond technology

Technology features in 62.7% of episodes that discuss a covered company. Consumer Cyclical, Communication Services, and Financial Services each appear in more than a third. The archive also contains substantial discussion of industrial businesses, healthcare, energy, and other sectors.

Shares use episodes discussing a covered company as their denominator. An episode discussing companies from several sectors counts in each relevant sector, so the shares do not sum to 100%.

Financial Services is particularly visible in executive appearances: 6,606 episodes compared with technology's 8,371. Within financial services, capital markets accounts for 1,576 executive episodes and asset management for 1,309. Elsewhere, auto manufacturers account for 1,219 and aerospace and defense for 952.

For an analyst, the useful search may begin with the business problem rather than a familiar podcast. A question about bank funding, industrial demand, or drug development can lead to a different set of shows and company perspectives. Sector and industry filters make that search manageable.

05 — Sentiment: find the months worth investigating

The monthly stance series tracks the balance between bullish and bearish claims across all subject types. Its lowest reading from January 2024 through August 2026 is 38.9% bullish in April 2025. The next lowest is 42.3% in March 2026.

Here, bullish share means bullish claims divided by bullish plus bearish claims. Neutral and mixed claims are excluded. This differs from the role comparison above, where percentages use all claims within each tier.

A change in this series gives a researcher a period to examine. Which companies or themes contributed? Did the arguments change, or did the mix of covered episodes change? Looking at the underlying claims can help distinguish those explanations.

The study does not test market correlation, predictive value, or investment returns. Its finding is a measurable change in the annotated conversation.

06 — Shows: choose breadth, depth, or both

Shows differ in both the frequency of executive appearances and the number of companies represented. Mad Money has 946 executive episodes spanning 381 companies. In Good Company with Nicolai Tangen spans 105 companies in 144 executive episodes. Leadership Next spans 107 in 123.

These rates describe the annotated episodes in this study, not every episode a show has published. They measure appearances in episode records, including any repeated or republished material captured by the sample, rather than verified unique interviews.

For broad discovery, begin with shows covering many companies. For a question about one business, search that company's appearances across shows and dates. The right reading list depends on the question you are trying to answer.

Start with a company you know

Bring a question from your coverage. Find the relevant appearances, compare the arguments, and follow them back to the source. The value of a large podcast archive is being able to do that work without deciding in advance which show will have the answer.

Explore podcast intelligence or see the annotation data.

Methodology and scope

The research team computed this analysis on September 15, 2026, from Matterfact's podcast_annotations collection, using one document per annotated episode. The sample covers December 4, 2006–September 15, 2026. It is an annotated sample of the wider archive, not a claim about the entire podcast market.

Periods. All-time tables retain the full supplied period. The annual view starts in 2021, and 2026 is labeled year to date. Monthly exhibits start in January 2024 and exclude incomplete September 2026. The recent monthly averages use the 20 complete months from January 2025 through August 2026. Historical counts reflect transcript availability.

Executive appearances. Current-officer counts follow the detailed source tables' insider_current definition, resolving roles as of each episode's air date. Board members and former officers are separate tiers. An appearance counts an episode, not a unique person, interview, or new statement. The covered-company set includes some private issuers.

Company discussion. A company is a subject when an episode contains sustained discussion of it, rather than a passing mention. Sector episode counts overlap when several sectors are discussed. Executive-appearance and discussion counts are separate measures; their ratio does not establish an exact overlap between the two episode sets.

Claims and attribution. The claims cover company, industry, macro, and commodity subjects. The role comparison describes a narrower subset of company claims. Role labels indicate the strongest participant tier present for a company on an episode, not verified sentence-level speakers. Quotes require transcript verification.

Stance and specificity. Role percentages include neutral and mixed claims in their denominator. Monthly and sector bullish share excludes them, using bullish / (bullish + bearish). Specific claims contain a quantity or a named metric, period, or source; this is separate from stance.

Excluded comparisons. The source describes a 2,000-company knowledge graph, while its sector and market-cap universe tables both total 2,994 companies. We omit universe-based representation percentages pending reconciliation. Sector claim totals also differ from the overall company-claim count, so they are not combined into an aggregate here. The analysis includes no market-return backtest or correlation estimate.

The original extraction and analysis are documented by the research team in studies/exec_appearances/extract.py and studies/exec_appearances/analyze.py in the API research repository.