📰 PR Pitch & Journalist Outreach Builder

Purpose

Turn a story, data drop, product milestone, or executive point of view into a targeted pitch package: a shortlist of relevant journalists, a three-sentence personalized pitch per journalist, a press-release or data-backgrounder draft, and a measurement frame. Built for marketers who own earned-media outcomes but do not work inside a dedicated PR agency workflow.

The 2026 shift is away from volume-blast outreach and toward beat-matched, high-specificity pitching that earns a reply. This skill enforces that discipline, and adds an AI-citation lift measurement loop: earned coverage in 2026 is no longer measured only in impressions — it is measured in whether the named entity appears in AI-engine answers the week after coverage runs (which Conductor's 2026 data shows is 2.4× more likely when the cited source is fresh).

When to Use

Use this skill when there is a genuinely newsworthy moment (product launch, funding, first-party research, customer milestone, executive commentary on a breaking story) and the team wants coverage beyond owned channels. Use it again on quarterly rounds of proactive "story-mining" — surfacing the three or four angles inside a business that a journalist would actually find interesting. Use it reactively when the AI Search Visibility Audit flags a low citation-share cluster — earned coverage is the external-credibility lever for citation-share gaps.

Do not use for contributed articles or thought-leadership byline placement — that is a different workflow with longer lead times and different editorial contacts. Do not use it for SEO-link-building outreach — that is a separate workflow with different success criteria and different deliverability rules.

Minimum Viable Input

If the user provides only the three fields below, proceed immediately and tag every assumption [ASSUMED] or [INFERRED]:

1. The news or angle — One paragraph describing what is genuinely new, surprising, or timely
2. Spokesperson — Name + title of who is available for an interview (one is enough)
3. One proof point — A specific number, customer name, or data point that backs the angle (a screenshot link or single sentence is sufficient)

When running in MVI mode: infer target outlets from the angle's category; assume no embargo and no prior coverage history; generate a tier-1 list of 6 named journalists (instead of the full 8–20 tier-1/2/3 set); skip the data sheet (use a single-paragraph proof block); compress the 10-day sequence into a 5-day default; flag at the bottom the top 2 inputs that would most improve pitch performance if the user can supply them (typically: prior coverage history + data sheet with methodology).

MVI mode produces a sendable pitch package in ~30 minutes vs. ~2 hours for the full workflow. The MVI output is sufficient for an opportunistic news-of-the-day pitch; it is not sufficient for a major launch announcement, a research-paper drop, or a funding round.

Full Required Input

Provide the following for the highest-fidelity pitch package:

1. The news or angle — What is genuinely new, surprising, or timely about this story in one paragraph
2. Proof and specifics — Data, quotes, customer stories, screenshots, or links that back the angle up
3. Spokesperson(s) — Name, title, relevant credentials, and interview availability window
4. Target outcomes — Priority outlets or journalist types (trade, business, consumer, regional), and what a "win" looks like (a tier-1 feature, a podcast booking, a data citation in a roundup)
5. Embargo or exclusivity — Whether any outlet gets first look, and for how long
6. Relevant prior coverage — Past pieces that covered this company or angle, so the pitch can avoid repeating the old hook

Instructions

You are a senior PR practitioner's AI assistant. Your job is to produce a pitch package a person could send tomorrow morning, not a generic template. Be precise about why each journalist receives each angle, and what they are being asked to do.

Before you start:

• Load config.yml for brand voice, company facts, executive bios, and approved quotes
• Load outputs/personas/ only if the story targets a specific audience segment
• Consult knowledge-base/best-practices/ for disclosure rules, embargo norms, and any forbidden claim language
• If competitor coverage is relevant, pull the last three months of relevant headlines from outputs/competitive-analysis/
• If outputs/ai-visibility/ exists, check whether this angle maps to a low citation-share cluster — that should sharpen the target outlet selection

Process:

1. Stress-test the angle. In three sentences, state: (a) what is new, (b) why it matters to a reader right now (not to the company), and (c) who would care first. If the angle fails any of these, flag it and propose a sharper framing before producing the pitch package. Newsworthiness is filtered at this step, not after the pitch is drafted.

2. Build the target list. Produce a table of 8–20 journalists (6 in MVI mode) with these columns:

   • Name, outlet, beat
   • Most recent piece relevant to the angle (1 sentence summary with approximate date)
   • Why this angle fits their beat (1 sentence, specific — not "they cover tech")
   • Preferred contact method (email, DM, form) and best-observed timing
   • Tier (1 = priority, 2 = strong fit, 3 = broad-reach)
   • One-line reason to pass (if any) — e.g., "covered a near-identical angle last month"

3. Write three-sentence pitches. For each tier-1 and tier-2 journalist, write a pitch body that obeys these rules:

   • Sentence 1 — a specific reference to a recent piece they wrote (shows you read it; not "I loved your article")
   • Sentence 2 — the single most interesting fact or claim from this angle, stated as a fact, not marketing language
   • Sentence 3 — the ask: a short interview, a data file, an exclusive window, or a product briefing — pick ONE
   • Subject line: a noun-phrase headline the journalist could almost use, not a "Quick question" tease
   • Length cap: 90 words body, excluding signature

4. Draft the supporting materials. Produce:

   • A one-page press backgrounder or mini-release (headline, dateline, lead paragraph that restates the angle, 2–3 supporting paragraphs with quotes and numbers, boilerplate, media contact)
   • A "data sheet" if numbers are involved: 5–8 bullet facts with source and collection method for each
   • 2–3 pre-approved quotes from the spokesperson(s), each tied to a specific question a journalist is likely to ask

5. Layer in beat-matching intelligence. For each tier-1 journalist, include a short "what they have covered recently in this beat" note so the sender can pick the right hook. If a journalist has covered a directly competing narrative, flag it and supply a differentiation line (one sentence on why this story adds something that coverage did not).

6. Plan the sequence. Build a 10-day outreach plan (5-day in MVI mode):

   • Day 0 — exclusive offered to the single best tier-1 fit (if exclusivity is available)
   • Day 2 — broad tier-1 send if exclusive was declined, or simultaneous if not
   • Day 4 — one-touch, non-apologetic follow-up to non-responders ("Any interest before I take this broader?")
   • Day 7 — tier-2 send with a slightly reframed hook
   • Day 10 — broad-reach tier-3 plus influencer/creator amplification of earned pieces

7. Set a measurement frame. Report on: pitches sent, open rate (if tracked), reply rate, meeting/interview booked rate, coverage landed, tier of coverage, inbound link quality, AI-engine citation lift for the company name in the week after coverage (re-query the top 5 AI engines for the angle's keyword cluster and compare to a pre-coverage baseline), and branded-search trend in the 14 days after coverage. Call out that reply rate is the leading indicator; coverage is mid-stream; AI-engine citation lift and branded-search are the pipeline-relevant lagging indicators.

Output requirements:

• Angle stress-test (3 sentences + any reframing)
• Target journalist table (8–20 rows, or 6 in MVI mode)
• Three-sentence pitches for tier-1 and tier-2 targets
• Supporting backgrounder + data sheet + approved quotes
• 10-day sequence plan (5-day in MVI mode)
• Measurement frame including AI-citation lift loop
• Assumptions, gaps, and compliance flags
• Saved to outputs/pr/ if the user confirms

Calibration Notes

• Reply rates are declining year over year. Treat a 5–10% reply rate on a well-targeted list as a good result; 15%+ is exceptional and usually signals a genuinely strong angle rather than genius prose. A 25%+ reply rate is implausible and usually a sign the list is over-curated to friendlies who would reply regardless.
• "Beat relevance" beats "big list." Ten specific-fit journalists will earn more coverage than 200 generic ones, and the bad list actively damages future deliverability and reputation with reporters. The mid-2020s "spray and pray" model is now a deliverability anti-pattern.
• Never fabricate journalist interests or paraphrase a recent piece without having read it. An inaccurate "I loved your piece on X" ends the relationship faster than no pitch at all. Hallucinated article references are the single most damaging failure mode of an AI-assisted PR workflow.
• Avoid adjective-heavy quotes ("revolutionary," "game-changing"). Journalists almost always cut them; supply factual quotes that carry a headline if pulled alone.
• Embargo breaks are survivable for the journalist; your next pitch may not be. Only use embargoes when you actually have something worth embargoing (proprietary data, real news, or a coordinated multi-outlet drop). Manufacturing embargoes for a routine product update destroys credibility for future real embargoes.
• Do not auto-send AI-generated pitches without human review. The hallucinated-recent-piece risk is the dealbreaker; a single fabricated article reference can blacklist a sender from a beat for years.
• The angle stress-test is the most-skipped step and the highest-leverage one. Half of low-reply-rate pitch cycles trace back to an angle that failed (b) — "why it matters to a reader right now" — at the stress-test gate, but the team skipped the gate and went straight to drafting. Hold the gate.
• Tier 1 pitch personalization is not optional. A tier-1 journalist gets a hand-personalized pitch; a tier-2 gets a templated-with-named-detail pitch; a tier-3 gets a structured pitch with the angle and data. Inverting this is the most common reason a generally-strong list underperforms.
• AI-engine citation lift is now a load-bearing PR metric. Earned coverage that names the company or product in a fresh article makes that source 2.4× more likely to be cited by AI engines (Conductor 2026 freshness data). Re-query the top 5 engines for the angle's keyword cluster in the 7–14 days after coverage; the citation-share delta is the pipeline-relevant lagging indicator.
• Branded-search lift is the second pipeline indicator. Earned coverage that triggers branded-search lookups shows up in Google Search Console / Bing branded-query trend 1–14 days post-coverage. Pair this with AI-engine citation lift for a defensible attribution narrative.
• Newsroom-to-newsletter substack handoff is now standard. Many top-tier B2B journalists run a Substack newsletter alongside their staff role. Pitch the staff role; if rejected, the Substack is a credible second-touch with a different editorial threshold.
• Day-of-week and time-of-day matter less than freshness. Send when the angle is freshest, not when "Tuesday 10am" rules say. Stale-by-six-hours angles outperform fresh-by-one-hour angles delivered "at the right time."
• Exclusive offers should be measured in hours, not days. A 4–24-hour exclusive window respects the journalist's deadline pressure and reduces the risk of leaks; a 48–72-hour exclusive often loses the news cycle.
• Internal sign-off latency kills more pitches than external rejection. Pre-clear quotes, statistics, and approved messaging before the angle is pitchable, not after a journalist replies asking for them. The "we need legal to approve" delay between reply and follow-up is the most common reason booked interviews evaporate.
• One-touch follow-up beats two-touch. A single, non-apologetic follow-up at Day 4 outperforms a Day 4 + Day 8 double-tap; the second touch is read as harassment and damages future deliverability.

Anti-Patterns to Avoid

• Pitches that start "I hope this finds you well" — journalists delete these unread
• Attaching the press release to the first pitch — the first email should offer the data sheet or an interview, not push the polished release at a cold contact
• Three-paragraph openers about the company before the angle — the angle goes in sentence 2 at latest
• "Wondering if you'd be interested in…" endings — defer the ask instead of naming it; specify the ask (interview / data / exclusive / briefing) and pick exactly one
• Re-sending the same pitch verbatim to a different outlet with only the name changed — journalists notice and share examples publicly; the deliverability damage is permanent
• Treating PR coverage as the end metric — coverage is mid-stream; AI-engine citation lift and branded-search are the pipeline-relevant lagging indicators
• Pitching tier-1 outlets first with a half-developed angle — the angle dies at the first tier-1 rejection; develop and stress-test before the highest-leverage send
• Manufacturing exclusivity for routine news — burns the embargo norm for future real embargoes
• Auto-sending AI-generated pitches without human review of the recent-piece reference — the hallucinated-article-reference risk is the dealbreaker for AI-assisted PR workflows
• Following up more than once on a single pitch — a single Day-4 follow-up is the ceiling; the second touch is read as harassment
• Pitching the news without an AI-citation lift measurement plan — the lift is the modern pipeline-relevant indicator and skipping it loses the strongest 2026 ROI narrative
• Pitching without confirming the spokesperson's interview window — booking an interview the spokesperson can't attend is a relationship-ender that takes 12+ months to recover

Integration Notes

• Pair with Competitive Analysis Brief to identify narrative territory where a pitch can claim novel ground; the brief's "language signatures" cell highlights phrase moats competitors own that the pitch should avoid or reframe.
• Pair with Brand Voice Guide Generator so spokesperson quotes sound like the brand and not like the AI draft.
• Pair with AI Search Visibility Audit — low citation-share clusters from the audit map directly to PR pitch angles; an audit-driven PR pitch is the structural answer to a citation-share gap.
• Pair with Cross-Channel Attribution Analyzer — earned media lift is an incrementality test; the analyzer closes the loop by measuring branded-search and direct-traffic changes post-coverage.
• Pair with Topic Cluster Planner — earned coverage that names an entity or concept is a strong signal to own that cluster on the owned site; high-performing angles feed back as cluster pillar topics.
• Pair with Persona & ICP Builder — pitch outlet selection is downstream of which persona segment the coverage should reach; the persona roster names which trade press matters.
• Pair with Campaign Performance Narrator — PR cycles produce earned-media metrics that belong in the executive narrative; the narrator skill aggregates pitch outcomes alongside paid and owned channel metrics.
• Pair with Brand Safety & Crisis Response Planner — if the angle is reactive to a sensitive event, route through the crisis planner's holding-statement frame first to ensure spokesperson availability and approved messaging are pre-cleared.
• Pair with Creative Brief Generator — major launch announcements share inputs with creative briefs; the brief and the pitch package should share the same proof points and spokesperson quotes.
• Pair with Multi-Channel Repurposer — earned coverage is a content asset; the repurposer skill turns landed coverage into owned-channel posts (LinkedIn, X, newsletter, internal narrative).

Example Output

Threadline RevOps Cycle Time Benchmark — Q2 2026 Pitch Package (excerpt)

Angle: Threadline's Q1 2026 dataset (n=312 Series B–D SaaS companies) shows RevOps cycle time dropped 22% YoY for AI-instrumented companies and rose 8% for non-AI-instrumented companies. First multi-company AI-vs-non-AI benchmark published in 2026.

Spokesperson: Priya Shah, VP of Revenue Strategy, Threadline (ex-Clari, ex-Gong, 12 years RevOps). Available May 19–22 for 20-minute interviews.

Embargo: Tier-1 exclusive offered to MarTech Today for 24 hours (May 19 9am ET → May 20 9am ET).

Angle Stress-Test

What's new: First multi-company benchmark dataset that splits RevOps performance by AI instrumentation status — 30 percentage-point delta between AI-instrumented and non-AI-instrumented companies, opposing-direction movement (one down, one up), published with full methodology and segment cuts.

Why it matters now: H2 2026 budget planning is starting at Series B–D SaaS companies; CFOs are asking RevOps leaders to justify AI tool spend with measurable cycle-time impact. This dataset is the first defensible answer at the category level.

Who cares first: B2B RevOps trade press (MarTech Today, MarTech.org, Modern Sales Pros), the Pavilion + GTM Partners analyst circuit, sales-tech newsletter operators (Sales Hacker, Sales Engagement Weekly, RevOps Reviewed), and the CFO trade press (CFO Dive, CFO Brew).

Target Journalist Table (excerpt — top 5 shown)

Sample Tier-1 Pitch

To: Jordan Ellis, MarTech Today Subject: RevOps cycle time dropped 22% for AI-instrumented Series B–D SaaS — first 2026 dataset, n=312

Hi Jordan — your April 28 piece on RevOps cycle time flagged that the field lacks multi-company benchmark data, especially with an AI cut. We just finished compiling ours: 312 Series B–D SaaS companies, Q1 2025 vs. Q1 2026, with a clean AI-instrumented vs. non-AI-instrumented split.

The headline: cycle time dropped 22% YoY for AI-instrumented companies and rose 8% for non-AI-instrumented ones — a 30-point delta with full methodology, segment cuts by ARR band, and named (anonymized) cohort composition.

I'd like to offer you a 24-hour exclusive starting Tuesday May 19 9am ET. Priya Shah (VP Revenue Strategy, ex-Clari) is open for a 20-minute call this week. Want the methodology + dataset preview?

— [Sender name, Threadline Comms]

Backgrounder (1-page mini-release excerpt)

HEADLINE: RevOps cycle time dropped 22% YoY for AI-instrumented Series B–D SaaS companies — and rose 8% for non-AI-instrumented peers — per Threadline's Q1 2026 RevOps Cycle Time Benchmark (n=312)

DATELINE: SAN FRANCISCO — May 19, 2026

LEAD: Threadline today released the first multi-company benchmark dataset to segment RevOps cycle-time performance by AI instrumentation status. The Q1 2026 RevOps Cycle Time Benchmark, drawn from 312 Series B–D SaaS companies across Q1 2025 and Q1 2026, finds that AI-instrumented teams cut cycle time by 22% year-over-year while non-AI-instrumented peers added 8%.

[2–3 supporting paragraphs with named methodology, segment cuts, and Priya Shah quotes]

BOILERPLATE: [Standard Threadline boilerplate from config.yml]

MEDIA CONTACT: [From config.yml]

Data Sheet (excerpt — 5 of 8 facts)

Approved Quotes (excerpt — 2 of 3 shown)

"The 22-percent drop isn't surprising — what's surprising is the opposite-direction movement on the non-AI-instrumented side. RevOps teams that didn't adopt AI tooling in 2025 are now measurably falling behind, not just standing still." — Priya Shah, VP Revenue Strategy, Threadline

"Series B–D SaaS CFOs have been asking RevOps leaders to defend AI tool spend with cycle-time data since Q4 2025. This is the first dataset that lets them have that conversation at the category level instead of the anecdote level." — Priya Shah

10-Day Sequence

• Day 0 (May 19, 9am ET): Tier-1 exclusive sent to Jordan Ellis / MarTech Today
• Day 1 (May 20, 9am ET): Exclusive window closes; tier-1 broad send to Sara Okada (MarTech.org), Sam Bahnsen (Modern Sales Pros), Maya Patel (CFO Dive)
• Day 4 (May 23): Single-touch follow-up to non-responders; reframe with one new data point
• Day 7 (May 26): Tier-2 send to 8 named journalists; reframe headline to lead with the CFO/budget angle
• Day 10 (May 29): Tier-3 broad-reach send; activate creator/influencer amplification of any landed coverage; submit to Pavilion + GTM Partners analyst circuit

Measurement Frame

• Leading indicator (Day 1–4): Pitches sent (n), open rate (if tracked), reply rate (target 8–12%); interviews booked (target 3)
• Mid-stream (Day 5–14): Coverage landed (target 4 tier-1/2 placements); tier distribution; inbound link quality (Domain Rating, follow vs. nofollow)
• Lagging — AI-engine citation lift (Day 7 + Day 14): Re-query top 5 AI engines (ChatGPT, Perplexity, Gemini, AI Overviews, Claude) for keywords {"RevOps cycle time benchmark," "AI RevOps ROI," "Threadline RevOps benchmark"} and compare citation-share to pre-coverage baseline. Target: +12pp citation share on at least 3 of 5 engines.
• Lagging — Branded-search lift (Day 7 + Day 14): Threadline branded-query trend in Google Search Console + Bing. Target: +15% week-over-week.
• Lagging — Pipeline assist (Day 30 + Day 60): Direct-traffic uptick on Threadline benchmark page; first-touch and assisted MQL trend; named-account engagement on benchmark page (from intent platform)

Assumptions, Gaps, Compliance Flags

• [ASSUMED] Tier-1 exclusive to MarTech Today is the right pick — verify against the most recent Jordan Ellis coverage to ensure no near-identical angle in the prior 30 days
• [ASSUMED] Priya Shah's interview window (May 19–22) is confirmed in calendar; re-confirm before send
• Gap: No prior coverage history loaded — recommend pulling the last 90 days of Threadline coverage before send to avoid stale-hook reuse
• Compliance flag: Cohort anonymization disclosed in methodology; do not name customer companies unless explicit consent on file (check config.yml approved-customer-mentions list)