Version Compatibility

Reference framework authored 2026-06 from a synthesis of public sources. No code dependencies. Update when:

• A peer-reviewed update to the AAAS Toolkit's podcast guidelines is published
• Anderson's TED principles are revised in a subsequent publication
• The hosting platform (Spotify, Apple, etc.) changes its episode format and metadata fields

If episode length, audience profile, or format (interview vs. narrated story vs. panel) differs from the defaults, scale the act structure; do not change the one-idea hierarchy.

Podcast

A scientific podcast is the long-form spoken version of the same research story the conference talk, poster, and press release are trying to tell. Where the conference talk has 12 minutes and the press release has 400 words, the podcast has 20-60 minutes. The longer format is a feature: it allows the host to do the work the conference talk cannot — to land the result, explain the mechanism, address the limitations, and end on the implication. The longer format is also a risk: a 45-minute episode with no structure is 45 minutes of wandering.

This skill encodes the structural moves that make a science podcast an effective long-form communication, not a slow news report.

The Three Format Modes

A scientific podcast has three production modes. Choose the mode by content type, not by production convenience.

A 12-minute research talk is a 12-minute conference talk; a 30-minute interview is a podcast interview. The format follows the content.

The One-Idea Transmission

Anderson's TED principles, applied to science communication, are: a great talk is the transmission of one idea. The podcast is no different. The one idea is the take-home message; the podcast is the argument that the take-home message is true.

The Take-Home Test

Write the take-home message in one sentence a non-specialist could repeat to a colleague. Test against three rules.

1. It is a claim, not a topic. "We studied mRNA methylation" is a topic. "Blocking METTL3 reverses temozolomide resistance in glioblastoma" is a claim.
2. It has a verb that asserts. "Show," "demonstrate," "cause," "predict," "explain."
3. It can be falsified. If a listener could disagree, the message is sharp enough to defend.

The take-home message is the contract with the listener. If the podcast cannot defend the message, it has not been a podcast; it has been a tour.

The One-Listener Test

The podcast is addressed to one person: the listener who knows nothing about the topic. The host's job is to speak to that listener, not to the guest, not to the host's peers, not to the field. Every technical term is defined as if the listener is hearing it for the first time.

The Three-Act Structure

A 30-45 minute episode has three acts, each with a specific function. The acts are not "intro, body, conclusion"; they are "why this matters, what we found, what it changes." The act boundary is signposted; the listener's mental model of the episode is the three-act structure.

The 18 minutes in Act 2 (for a 30-minute episode) is the meat of the episode. The temptation is to spend all 30 minutes on Act 2 because that's where the science is. The rule is that Acts 1 and 3 are necessary: Act 1 is the hook that keeps the listener, and Act 3 is the implication that gives the listener a reason to remember. A podcast that is all Act 2 is a lecture.

The Act 1: Why This Matters

Act 1 has three moves.

1. The Hook (30-60 seconds). A story, a question, a surprising fact. The hook is not "today we're talking to Dr. X about Y." The hook is a 30-second anecdote that captures the listener's attention.
2. The Problem (2-3 minutes). Why the topic matters now. A specific example: a patient, a failed trial, a public health crisis. The problem is not the technique; it is the human or societal need.
3. The Guest (1-2 minutes). Who the host is talking to and why the guest is the right person to talk to. The introduction establishes credibility.

Hook Examples

"When Jane Doe was diagnosed with glioblastoma in 2018, her doctors gave her 15 months. The chemotherapy worked for six months, and then the cancer came back. Her doctors had no second option. Today, we're going to talk about a new discovery that might change that."

This hook is a story. The listener hears Jane's situation, understands the problem, and wants to know whether the new discovery would have helped.

Bad: "Today we're talking to Dr. Smith about RNA modifications in glioblastoma."

This hook is a topic. The listener hears an acronym and a topic. The listener has no reason to keep listening.

The Act 2: What We Found

Act 2 is the deciding experiment and its mechanism. The act is the core of the podcast; it is where the listener learns the result. The act has three sub-moves.

1. The Approach (3-5 minutes). How the researchers did the work, in plain language. Define the technique using a metaphor.
2. The Deciding Experiment (8-12 minutes). The one result that, if it failed, the paper fails. This sub-move is the spine of the episode.
3. The Validation (3-5 minutes). The supporting evidence that makes the result credible.

Approach as Metaphor

Every technique in a podcast is described twice: once as a metaphor, once as the technical term. The metaphor is for the listener; the technical term is for the record.

"The technique we used is called m6A-sequencing, but you can think of it as a highlighter. We're highlighting every place in the cell's RNA that gets a chemical mark, the m6A mark, so we can see which genes are tagged."

The metaphor does the work of the technical term. The listener understands the technique; the technical term is on the record.

The Deciding Experiment

The deciding experiment is the one result that anchors the paper. In a podcast, the deciding experiment gets 8-12 minutes of the 30-minute episode. The rule is the same as the conference talk: if a hostile reviewer could attack only one figure, which figure would they attack? That figure is the deciding experiment.

In a podcast, the deciding experiment is described in three layers.

1. The setup (1-2 minutes): What the researchers did.
2. The result (1-2 minutes): What they found.
3. The interpretation (3-5 minutes): What the result means, why it is the deciding experiment.

The interpretation is the work. The host's job is to ask "what does this mean?" and "why is this the deciding experiment?" until the listener understands.

The Act 3: What It Changes

Act 3 has three moves.

1. The Implication (2-3 minutes). What the result changes for the field, for patients, for society. The implication is concrete, not abstract.
2. The Limitation (2-3 minutes). What the result does not show, what is still unknown. The limitation is honest, not a hedge.
3. The Close (1-2 minutes). A restatement of the take-home message, a thank-you to the guest, a call-to-action for the listener.

Limitation as Honesty, Not Hedge

The limitation is not "we hope to learn more." The limitation is a specific gap: the study was in mice, the inhibitor is preclinical, the cohort was 12 patients, the follow-up was 6 months. The listener needs to know the limits to evaluate the result.

"Before we close, I want to flag the limitations. This study was in mice. We don't know if the same mechanism applies in humans. The METTL3 inhibitor is a preclinical compound; it's not yet in clinical trials for glioblastoma. So the result is promising, but it's not yet a therapy."

The limitation is the contract with the listener: we are telling you what we know and what we do not know. The host is not selling; the host is informing.

The Host Preparation

A podcast is only as good as the host's preparation. Apply four preparation rules.

Rule 1: Read the Paper Twice

The host reads the paper once for content, once for the deciding experiment. The first read establishes the territory. The second read identifies the spine: the one figure, the one result, the one sentence that the paper is built around. The second read is what makes the host's questions sharp.

Rule 2: Write 20 Questions, Use 5

The host writes 20 questions, ranked by priority. The top 5 are the spine of the episode; the next 5 are follow-ups; the next 5 are wildcards; the last 5 are for the closing. During the recording, the host uses the top 5 and follows the guest's lead on the next 10. The last 5 are backups.

Rule 3: Pre-Record the Cold Open

The cold open is the first 60-90 seconds. It is the hook, the moment that makes the listener keep listening. The cold open is pre-recorded and edited to perfection. It is not improvised.

Rule 4: Mark the Act Boundaries

The host marks the act boundaries during the recording. The transitions ("that's the setup; now the result"; "now the implication") are verbal signposts. The listener needs the signposts; the absence of signposts is why most podcasts feel like wandering.

The Guest Preparation

A guest who has not been prepared is a guest who will speak in jargon. Apply three preparation rules.

Rule 1: Send the Take-Home in Advance

The host sends the guest the take-home message in writing 48 hours before the recording. The take-home is one sentence. The guest knows what message the episode will land.

Rule 2: Send a Glossary

The host sends the guest a glossary of technical terms that will be used in the episode. The guest knows which terms to define and which to use directly.

Rule 3: Set the "No Jargon" Rule

The host tells the guest: "If you use a term I don't know, I will stop you and ask you to define it. This is the format." The guest is then licensed to be stopped, which gives the listener a clearer episode.

The Show Note

A show note is the 200-400 word description of the episode that lives on the podcast platform. The show note is not a transcript; it is a sales document. Apply four rules.

1. First sentence is the take-home message. "In this episode, we talk to Dr. Jane Doe about her discovery that blocking METTL3 reverses temozolomide resistance in glioblastoma." The first sentence is the soundbite.
2. Second sentence is the implication. "We discuss the implications for combination therapy and what comes next for clinical trials."
3. Third sentence is the guest bio. "Dr. Doe is a professor of oncology at University Hospital and the lead author of the paper published in Nature."
4. Fourth and following sentences are timestamps and links. "00:00 - Why this matters. 05:00 - The deciding experiment. 30:00 - Implications. Links to the paper and the guest's lab."

The show note is the document that converts a listener's curiosity into a download. A show note that buries the take-home message in the third paragraph has failed.

The Transcript

A podcast transcript is the searchable text version of the episode. The transcript is not a verbatim copy; it is an edited version that fixes filler words ("um," "uh"), removes false starts, and tightens the language. The transcript is what a listener reads when they cannot listen, and what a search engine indexes.

The transcript is the document that the search engine sees, the document that the listener skims, and the document that a journalist uses to write a story. A podcast without a transcript is a podcast that is invisible to search and inaccessible to deaf listeners.

The Listener-Stress-Position Principle

Gopen & Swan's stress-position principle applies to the spoken register. The end of a sentence is the listener's stress position. The host's job is to design the sentences so the stress position carries the new information.

Bad: "When researchers blocked METTL3, the chemotherapy temozolomide became effective again." Good: "When researchers blocked METTL3, the chemotherapy temozolomide became effective again."

In the bad version, the new information is "became effective again" — and it is at the end. The good version places the same new information at the end, with a 1-second pause before it to mark the stress.

The host applies the principle in two ways.

1. Sentence design. Place the new information at the end of the sentence.
2. Pause discipline. A 1-second pause before the new information marks the stress position and gives the listener time to process.

The pause is the spoken equivalent of italics. The host's silence is the listener's processing time.

The Anderson TED Principle

Anderson's TED principles for science talks translate to podcasts in four moves.

1. The new is more memorable than the known. The host should spend more time on what the researchers found than on what was already known. The new is the news; the known is the context.
2. The frame is more memorable than the fact. A result framed as a story is more memorable than a result framed as a fact. The host should frame the result as a story: who tried what, who found what, who changes what.
3. The single idea is the contract. A podcast that tries to land three ideas lands none. The host should commit to the take-home message and reinforce it three times: in the cold open, in the deciding experiment, in the close.
4. The 18-minute attention span is the limit. A podcast that runs longer than 45 minutes risks losing the listener. The host should edit for length; the show note is the document for the listener who wants more.

The Common Failure Modes

A 30-Minute Worked Example

For a 30-minute interview on "METTL3 inhibition reverses temozolomide resistance in glioblastoma," the structure is:

• 0:00-0:30 Cold open. Jane Doe's story. 15-month survival, no options.
• 0:30-2:00 Act 1 problem. Glioblastoma: 15-month median; resistance is the bottleneck.
• 2:00-5:00 Act 1 guest. Dr. Smith is a glioblastoma researcher at University Hospital; published in Nature 2024.
• 5:00-7:00 Act 2 approach. m6A-sequencing, defined as a "highlighter" technique.
• 7:00-15:00 Act 2 deciding experiment. The METTL3 knockdown / overexpression experiment; the m6A site on MYC mRNA; the in vivo rescue.
• 15:00-22:00 Act 2 validation. The patient samples, the dose response, the rescue in xenografts.
• 22:00-25:00 Act 3 implication. Combination therapy; clinical trials in 2026; what this changes for patients.
• 25:00-28:00 Act 3 limitation. Mouse study; preclinical inhibitor; needs human data.
• 28:00-30:00 Close. Restate the take-home; thank the guest; call-to-action for the listener.

Total: 30 minutes. The deciding experiment (7:00-15:00) is 8 minutes — the longest single segment. The implication (22:00-25:00) is 3 minutes. The cold open is 30 seconds. The show note is 200 words with the take-home message in the first sentence.

Production Checklist

References

• Anderson, The Way We're Working Isn't Working — TED principles: new over known, frame over fact, single idea, 18-minute attention span.
• Gopen & Swan, "The Science of Scientific Writing" — stress position, applied to spoken register.
• AAAS Communication Toolkit — audience-first discipline, plain language.
• NIH Plain Language guidelines — jargon discipline, quantified claims, active voice.
• Alda Center for Communicating Science — improvisation, clarity, present-moment speaking.
• Nature Masterclasses on Podcast Production — pre-production, show note, transcript.

Related Skills

• scientific-communication/conference-talk — short-form spoken version of the same story.
• scientific-communication/poster — visual version for conferences.
• scientific-communication/elevator-pitch — distilled version for the hallway conversation.
• scientific-communication/press-release — written version for journalists and the public.