Podcast intelligence: strategy stated in a competitor's own words

A competitor’s website is written by committee, reviewed by legal, and optimised. It tells you what they want to be understood as.

A forty-minute podcast interview with their VP of Operations is a person talking, unscripted, to a friendly host, about what they are actually working on.

The second is worth considerably more, and it is public.

Why executive interviews leak strategy

The format works against guardedness in three ways.

It is long. Forty minutes is enough time to run out of prepared answers. The interesting material is usually past the twenty minute mark, after the potted history and the origin story.

The host is friendly. Industry podcasts are not adversarial journalism. Guests relax, and relaxed people explain their reasoning rather than reciting positions.

Nobody treats it as marketing collateral. Website copy gets reviewed. A podcast appearance gets a calendar invite. The asymmetry in scrutiny is enormous and it is where the value lives.

The result is that executives routinely describe, in plain terms, what they are building, what markets they are moving into, what they think is wrong with the category, and what they are worried about. Months before any of that reaches a website.

What to search for

The lane is worth running because transcripts make it searchable at scale. Pull episodes, pull transcripts, and search across the whole corpus.

Brand mentions across the category. Not just your competitors’ own appearances. Where is your category being discussed, and who is being named as an example of doing it well? That is third-party positioning evidence and it is more credible than anything a company says about itself.

Forward-looking language. “We are starting to”, “next year we will”, “we have been investing in”. A statement about future intent, on the record, with a date. That is a leading indicator with a timestamp on it.

Complaints about the category. Executives are unusually candid about what frustrates them in an industry, and those complaints are usually the seeds of what they intend to fix. A competitor saying that customers hate a particular part of the process is describing their own roadmap.

Who is buying podcast advertising. This is the underused half. Ad reads are a media buy, and they tell you who considers this audience worth paying for. Sometimes the advertiser is a competitor. More often it is an adjacent company whose presence tells you something about how the category is being served.

The reads themselves are worth capturing verbatim, because ad copy written for audio is unusually plain about the value proposition. Nobody writes a podcast read in the abstract corporate register.

The dating discipline

Every transcript carries a publication date, which turns a corpus into a chronology.

A statement of intent from eighteen months ago that never materialised is a different fact from the same statement made last month. The first tells you about a plan that stalled, which is genuinely useful. The second tells you about something in flight.

The same applies to absence. A company that appeared on four podcasts in 2023 and none since has either become less interesting to hosts or stopped prioritising it. Combined with dormant content and no new campaigns, that starts to describe a company in maintenance mode.

The lane we did not run

Here is the part worth being straight about, because it says something about how these lanes should be billed.

We did not run the podcast lane on the first live deployment.

That client’s category has effectively no podcast presence. Running the lane would have spent real money to confirm silence. So it was not run, and the report said so rather than presenting a thin result as a finding.

Lanes are chosen per market, not billed as a checklist. A competitive audit that runs the same twelve lanes on every engagement is selling process rather than findings, and the client pays for research into surfaces their category does not use.

The judgement about which lanes to run is part of the work. In a category with a strong podcast culture, professional services, technology, some parts of healthcare, it is among the highest-signal lanes available. In a category where nobody records anything, it is a line item.

Verification still applies

Transcripts create a specific temptation, and it is worth naming.

Auto-generated transcripts contain errors. Speaker attribution drifts. Homophones get chosen wrongly. Numbers get mangled.

So a quote pulled from an auto-transcript is not automatically evidence. It has to be verified against the audio at the timestamp before it becomes a record, and the record should say that it was.

Our practice is to note it explicitly:

Transcript is auto-generated. Wording verified against audio at 14:07, but surrounding context was not fully reviewed.

Confidence medium rather than high, with the caveat attached to the record where somebody deciding what to do with it will see it.

That is more work than copying a line out of a transcript. It is also the difference between a quote a client can put in a board deck and one that falls apart when a competitor listens to the episode.

Where it fits

Podcast findings feed primarily into Positioning, because a stated point of view in an executive’s own words is the strongest available evidence that a position is real rather than website decoration.

The tier 5 requirement for Positioning asks for evidence from outside a company’s own properties. A transcript from a third-party show, dated, verified, is exactly that.

Which is a useful thing to know in both directions. If you want to be scored well on a lens that requires third-party corroboration, being interviewed by people in your own category is one of the cheaper ways to earn it.

The full list of research lanes and how they are selected per market is on the competitive intelligence page.