--- name: prospect-research description: Turns a single LinkedIn URL into a complete pre-meeting prospect profile by mining the prospect's whole public footprint, including LinkedIn posts, podcasts, articles, talks, press, YouTube, and X, then synthesizing verbatim quotes, recurring themes, a company profile, and a 4-quadrant "Reasons to Layer" outreach ammunition stack. Use when the user says "research this prospect", "build me a pre-meeting profile", "what's the ammo on [name]", "dig up everything on this LinkedIn", or pastes a LinkedIn URL before a meeting. license: MIT metadata: author: Chris Merrill version: 2.0.0 persona: Strategic Steve rave: R4 A4 V3 E3 category: sales source: agentai-action --- # Prospect Research ## What it does One LinkedIn URL in, a deep, sourced dossier out. It reads the prospect's own words across their full public footprint (posts, podcasts, articles, talks, press, video), extracts verbatim quotes you can reference by name, and packages everything into a "Reasons to Layer" ammunition stack a rep can open a meeting with. Wider and slower than a quick brief, and worth it when the meeting matters. ## When to use - "Research this prospect before my call", "build a pre-meeting profile", "what's the ammo on [name]". - "Dig up everything on this LinkedIn", or a pasted LinkedIn profile URL ahead of a named meeting. - Not for bulk list scoring or quick lookups; this is a deep pass on one person. For a lighter profile without the quote mining, `prospect-intelligence-deep-dive` is faster. ## Inputs Required: the prospect's LinkedIn profile URL. Derived automatically: their name and company, parsed from the profile. Nothing else to fill in. ## The bar - Verbatim means verbatim. A quote appears exactly as said or written, with source, date, and URL, or it does not appear. - Every Reasons-to-Layer bullet is one line and carries its source. An unsourced bullet is a liability in a live meeting. - Themes are patterns across at least two artifacts, not one post stretched into a worldview. - Prefer omission over a guess, everywhere. The rep will say these things out loud to the prospect's face. ## Workflow 1. **Fetch LinkedIn.** Pull the profile and roughly the last 50 posts. Parse name, current title, career history. Rank the posts by substance: original opinions outrank reshares, specifics outrank platitudes. 2. **Map the public footprint.** Search the person and company across podcasts, articles, Substack, Medium, conference talks, press, YouTube, and X, plus recent company news. De-duplicate and drop anything not clearly this person. 3. **Extract quotes, verbatim.** Capture exact, attributable statements grouped by artifact (source type, title, publication, date, URL). Paraphrase nothing. Drop anything with murky attribution. 4. **Synthesize themes and personal signals.** Recurring stances held across artifacts, plus human details worth knowing (they coach youth soccer, they hate the word "synergy"). Two artifacts minimum per theme. 5. **Profile the company.** Size, stage, recent news, positioning, customer type. Enough to place the person's opinions in their business context. 6. **Write the About blurb.** Two or three plain sentences: current role and scope, career arc, what makes them distinctive. No adjectives you cannot back. 7. **Build Reasons to Layer.** The 4-quadrant stack, every bullet one sourced line: - Person: career trajectory, education, tenure, past companies, location. - Business: current scope, seniority, what they own, focus areas. - Social: sharpest opinions and short quote highlights, 15 words or fewer each. - Company: size, stage, recent news, positioning. 8. **Assemble the profile** in the output format, quotes and sources intact. ## Output ``` ## [Name] pre-meeting profile **About:** [2-3 sentence bio] **Company:** [name · size · stage · recent signal] ### Reasons to Layer - Person: [one-line fact] (source, url) - Business: [one-line fact] (source, url) - Social: "[quote of 15 words or fewer]" (source, url) - Company: [one-line fact] (source, url) ### Notable quotes - "[verbatim quote]" ([publication, date, url]) ### Themes and personal signals - [Recurring stance or human detail, with the artifacts behind it] ``` ## Worked example ``` Fictional example. Never reuse these names or numbers in real output. ## Tomas Herrera pre-meeting profile **About:** Tomas Herrera is COO of Fieldstone Payroll, a 60-person payroll provider for construction firms. He spent nine years at a national payroll company before joining Fieldstone in 2022, and writes often about why mid-market service companies underinvest in onboarding. **Company:** Fieldstone Payroll · ~60 employees · bootstrapped, profitable · launched a subcontractor compliance product in January ### Reasons to Layer - Person: 9 years at Paychex before Fieldstone, started as an implementation lead (LinkedIn profile) - Business: owns ops, support, and implementation; support response time is his stated 2026 focus (LinkedIn post, Jan 9) - Social: "Onboarding is the product. Everything after is retention." (Built to Serve podcast, Nov 14) - Company: compliance product launch doubled inbound demo requests, per their announcement (company blog, Jan 21) ### Notable quotes - "Onboarding is the product. Everything after is retention." (Built to Serve podcast, Nov 14, url) - "We lose deals to spreadsheets more than to competitors." (LinkedIn post, Oct 2, url) - "I'd rather hire a dispatcher than a dashboard." (Construction Ops Summit panel, Sep 18, url) ### Themes and personal signals - Onboarding as the core product: the podcast, two LinkedIn posts, and the summit panel all land the same stance. - Skeptical of tooling without service behind it; sells against "dashboard vendors" explicitly (podcast, panel). - Personal: runs a Saturday trades-mentorship program he mentions in two posts; genuine door-opener if it comes up naturally. ``` ## Never - Never paraphrase and present it as a quote. If the exact words are unavailable, summarize openly ("he has argued that...") without quote marks. - Never include a quote, fact, or personal detail without a source and, where the data provides one, a direct URL. - Never build a theme from a single artifact. - Never surface personal details that would feel invasive said aloud in a meeting (family, health, politics, religion). Mentorship programs yes; custody battles never. ## Recovery - Thin public footprint: build from the profile and company-level signals, flag lower confidence, and shrink the Social quadrant rather than filling it with reaches. - LinkedIn unreachable or private: ask for another handle or artifact; if you cannot ask, research name plus company across the open web and say the profile itself was unavailable. - Common name, ambiguous results: keep only artifacts where the company or role confirms identity, and say what you excluded. - Prolific writer, too much material: keep the sharpest 5 quotes and 3 themes ranked by recency and relevance to the meeting, and say the corpus was larger. ## Notes - Tool references (LinkedIn fetch, web scraping, search) are connector-agnostic; map them to whatever your environment exposes. Read-only: it produces a report, it writes nowhere.