Public Relations
Advertising is you talking about yourself. PR is arranging for someone else to. The entire value is in the someone else, which is why every shortcut that fakes the third party destroys the thing it was trying to produce.
The founding insight, and its correction. Bernays' contribution was the sequence: reach the group's trusted leaders and the group follows, rather than arguing with the crowd directly. That still works and most people still skip it. His method — undisclosed client-funded "independent" surveys and front groups — is now illegal, and his own account of his results was inflated, as his independent biographer documented. Take the sequence; refuse the concealment. The marketing-side application of that same doctrine is in the marketing skill's persuasion reference; this skill covers the practice of earning and defending the third-party voice itself.
The finding that should govern everything you do here
Ma & Zhan's meta-analysis (2016, Journal of Public Relations Research, across 24 studies) measured what actually drives reputation after a crisis:
- What you are blamed for → reputation: r = −.54 (strong)
- Which response strategy you choose → reputation: r = .23 (weak)
What you did matters roughly twice as much as anything you say about it. This is the most useful and least flattering finding in the field, and it should make any communications professional humble about their own leverage. The implication is structural: most reputation work is operational, not rhetorical. If the substance is bad, comms buys you very little, and the honest advice is to fix the substance.
Corollary from the same literature: Coombs & Holladay (2008) tested apology against sympathy against compensation. All three performed the same, and all three beat merely providing information. So the agonized "should we apologize" debate that consumes crisis meetings is mostly a false dilemma. What matters is that you respond accommodatively at all.
The procedure
- Check whether PR is the right instrument. It is not, if: there is no genuine news, the product isn't ready for scrutiny, nobody has bandwidth to respond to journalists, or the underlying problem is operational. A skill that always says yes is useless. Say so and stop.
- Find the actual story. Not your announcement — the story a journalist's readers would want. Funding rounds and product updates are announcements; a surprising number, a named conflict, a genuine first, or a human consequence is a story.
- Identify the trusted third parties for your specific audience, then earn them. Analysts, maintainers, practitioners, specific beat reporters. Bernays' sequence, applied honestly.
- Pitch, or place. → references/earned.md
- Measure outcomes, not clippings. → references/earned.md (measurement)
- When it goes wrong, follow the crisis protocol — and disclose before you are discovered. → references/crisis.md
Where to read
| The task |
Open |
| Crisis, bad press, a mistake becoming public, holding statements, spokesperson prep, apology decisions |
references/crisis.md |
| Getting coverage: what journalists actually want, pitching, releases, exclusives vs embargoes, newsjacking, channels, measurement |
references/earned.md |
| Tactical B2B/startup journalist pitching, HARO/Qwoted workflow, media lists |
coreyhaines31/marketingskills → public-relations (MIT, excellent, already exists — don't rebuild it) |
| Persuading the audience once you have their attention |
the marketing skill |
| Getting investors or donors to fund you |
the fundraising skill |
The laws
- The deed outranks the statement. See the meta-analysis above. Comms cannot outrun conduct.
- Steal thunder. Disclosing a problem yourself, before someone else finds it, produces measurably better credibility and reputation outcomes than being caught. The instinct to wait and hope is the single most expensive instinct in this field.
- Relevance beats relationships, and both beat volume. Journalists cite beat relevance as their top criterion and reject most pitches for lacking it. They receive fifty-plus pitches a week.
- Never fake the third party. Astroturfing, undisclosed paid reviews, sockpuppet Wikipedia editing, invented consensus. All illegal, all discoverable, all fatal to the mechanism you were using.
- A press release is not a strategy. Around 3% of wire-distributed releases get picked up. Targeted relevance works; spray-and-pray does not.
- AVE is not a measurement. Advertising Value Equivalency — valuing coverage at notional ad rates — is formally rejected by the industry's own measurement bodies. It treats a hit piece and a glowing profile as identical.
- Distrust the field's canonical stories, including Tylenol. The gold-standard case has been publicly disputed within the profession: the first recall was regional not national, the response took days not hours, and the retelling centers the corporation rather than the seven people who died.
Ethics (non-negotiable)
The line is disclosure and truth of the underlying claim, not the existence of persuasion. Organized persuasion in a mass society is unavoidable; concealment is a choice.
Specifically prohibited, with real enforcement behind each: concealed-sponsor front groups and astroturf campaigns; undisclosed paid reviews, testimonials, or influencer content (the FTC's rule effective October 2024 carries penalties above $50,000 per violation and explicitly covers AI-generated fakes); undisclosed paid Wikipedia editing (a Terms of Use violation since 2014, not merely a guideline); engineered "expert consensus" built from leading survey questions toward a predetermined result — the exact Bernays pattern, later industrialized by tobacco; and fabricated credentials in pitches.
Read the exposé literature — Holiday on how the media pipeline is gamed, Tye on Bernays, the Cardiff churnalism data, the Nayirah testimony — not as a playbook but as the reason these rules exist. The techniques that manipulated a war are the same techniques scaled down to a product launch.
1---2name: public-relations3description: Earn third-party credibility and protect reputation when it is under attack. Use when the user wants press or media coverage, is planning an announcement or launch, needs a press release or media strategy, is deciding between an exclusive and an embargo, wants thought-leadership or podcast placement, is handling a crisis or bad press, needs a holding statement or spokesperson prep, is asking whether to apologize and how fast, wants to measure PR, or is asking where the line sits between legitimate earned media and astroturfing. Grounded in the evidence — Coombs' crisis research, journalist surveys, the churnalism studies — with the field's folklore separated from its findings. Includes the exposé literature, because knowing how this is abused is how you avoid it.4---56# Public Relations78Advertising is you talking about yourself. PR is arranging for someone else to. The entire value is in the *someone else*, which is why every shortcut that fakes the third party destroys the thing it was trying to produce.910**The founding insight, and its correction.** Bernays' contribution was the sequence: reach the group's trusted leaders and the group follows, rather than arguing with the crowd directly. That still works and most people still skip it. His method — undisclosed client-funded "independent" surveys and front groups — is now illegal, and his own account of his results was inflated, as his independent biographer documented. Take the sequence; refuse the concealment. The marketing-side application of that same doctrine is in the `marketing` skill's [persuasion reference](../marketing/references/persuasion.md); this skill covers the practice of earning and defending the third-party voice itself.1112## The finding that should govern everything you do here1314Ma & Zhan's meta-analysis (2016, *Journal of Public Relations Research*, across 24 studies) measured what actually drives reputation after a crisis:1516- **What you are blamed for → reputation: r = −.54** (strong)17- **Which response strategy you choose → reputation: r = .23** (weak)1819**What you did matters roughly twice as much as anything you say about it.** This is the most useful and least flattering finding in the field, and it should make any communications professional humble about their own leverage. The implication is structural: most reputation work is operational, not rhetorical. If the substance is bad, comms buys you very little, and the honest advice is to fix the substance.2021Corollary from the same literature: Coombs & Holladay (2008) tested apology against sympathy against compensation. **All three performed the same**, and all three beat merely providing information. So the agonized "should we apologize" debate that consumes crisis meetings is mostly a false dilemma. What matters is that you respond accommodatively at all.2223## The procedure24251. **Check whether PR is the right instrument.** It is not, if: there is no genuine news, the product isn't ready for scrutiny, nobody has bandwidth to respond to journalists, or the underlying problem is operational. A skill that always says yes is useless. Say so and stop.262. **Find the actual story.** Not your announcement — the story a journalist's readers would want. Funding rounds and product updates are announcements; a surprising number, a named conflict, a genuine first, or a human consequence is a story.273. **Identify the trusted third parties** for your specific audience, then earn them. Analysts, maintainers, practitioners, specific beat reporters. Bernays' sequence, applied honestly.284. **Pitch, or place.** → [references/earned.md](references/earned.md)295. **Measure outcomes, not clippings.** → [references/earned.md](references/earned.md) (measurement)306. **When it goes wrong, follow the crisis protocol** — and disclose before you are discovered. → [references/crisis.md](references/crisis.md)3132## Where to read3334| The task | Open |35|---|---|36| Crisis, bad press, a mistake becoming public, holding statements, spokesperson prep, apology decisions | [references/crisis.md](references/crisis.md) |37| Getting coverage: what journalists actually want, pitching, releases, exclusives vs embargoes, newsjacking, channels, measurement | [references/earned.md](references/earned.md) |38| Tactical B2B/startup journalist pitching, HARO/Qwoted workflow, media lists | `coreyhaines31/marketingskills` → `public-relations` (MIT, excellent, already exists — don't rebuild it) |39| Persuading the audience once you have their attention | the `marketing` skill |40| Getting investors or donors to fund you | the `fundraising` skill |4142## The laws4344- **The deed outranks the statement.** See the meta-analysis above. Comms cannot outrun conduct.45- **Steal thunder.** Disclosing a problem yourself, before someone else finds it, produces measurably better credibility and reputation outcomes than being caught. The instinct to wait and hope is the single most expensive instinct in this field.46- **Relevance beats relationships, and both beat volume.** Journalists cite beat relevance as their top criterion and reject most pitches for lacking it. They receive fifty-plus pitches a week.47- **Never fake the third party.** Astroturfing, undisclosed paid reviews, sockpuppet Wikipedia editing, invented consensus. All illegal, all discoverable, all fatal to the mechanism you were using.48- **A press release is not a strategy.** Around 3% of wire-distributed releases get picked up. Targeted relevance works; spray-and-pray does not.49- **AVE is not a measurement.** Advertising Value Equivalency — valuing coverage at notional ad rates — is formally rejected by the industry's own measurement bodies. It treats a hit piece and a glowing profile as identical.50- **Distrust the field's canonical stories, including Tylenol.** The gold-standard case has been publicly disputed within the profession: the first recall was regional not national, the response took days not hours, and the retelling centers the corporation rather than the seven people who died.5152## Ethics (non-negotiable)5354The line is **disclosure and truth of the underlying claim**, not the existence of persuasion. Organized persuasion in a mass society is unavoidable; concealment is a choice.5556Specifically prohibited, with real enforcement behind each: concealed-sponsor front groups and astroturf campaigns; undisclosed paid reviews, testimonials, or influencer content (the FTC's rule effective October 2024 carries penalties above $50,000 per violation and explicitly covers AI-generated fakes); undisclosed paid Wikipedia editing (a Terms of Use violation since 2014, not merely a guideline); engineered "expert consensus" built from leading survey questions toward a predetermined result — the exact Bernays pattern, later industrialized by tobacco; and fabricated credentials in pitches.5758Read the exposé literature — Holiday on how the media pipeline is gamed, Tye on Bernays, the Cardiff churnalism data, the Nayirah testimony — not as a playbook but as the reason these rules exist. The techniques that manipulated a war are the same techniques scaled down to a product launch.
DY