# Spot culture red flags in leadership interviews > Import a target's exec interviews and ask pointed questions — how do they talk about attrition, customers, each other? - For: Acquirer / Angel - Category: Finance, Investing & Due Diligence - MCP tools: library_import_youtube_video, deep_context - HTML version: https://veridive.com/mcp/use-cases/leadership-red-flags ## The problem Culture is the asset that never appears in the data room. You can diligence the contracts, the code and the cohort curves — but the thing that determines whether the team you're buying still exists in eighteen months is how its leaders actually operate, and no target volunteers that. The management presentation is theater; the reference list is hand-picked. Yet leaders narrate their own culture constantly, in public, without noticing. How a CEO talks about departures ("we've upgraded the team" versus "we failed those people"). Whether credit flows downward or pools at the top. How customers get described when the interviewer isn't a customer. Whether co-founders mention each other with warmth, formality, or not at all. These tells accumulate across interviews — and the interviews are all on the record. ## How veridive MCP helps **library_import_youtube_video** builds the exhibit file: the target's leadership interviews, panels and podcast appearances imported into your veridive library, tagged by executive, transcripts and all. Deal-quiet and reversible — you're collecting public record, not tipping your interest. **deep_context** runs the behavioral read: scope questions across the collection and get answers **cited to the exact chunks**. "How does the CEO talk about people who left?" "When something failed, who takes the blame in the telling?" "How do the two founders describe each other?" Pattern-level questions, answered across many appearances at once — because a red flag in one interview is a bad day, but the same flag across two years of them is the culture you're about to buy. ## How to run it 1. **Connect veridive MCP** — One-time setup: add the veridive MCP server to Claude, ChatGPT, Cursor or any MCP client and generate your API key. Takes about two minutes. 2. **Build the exhibit file** — Import the leadership team's public appearances — CEO, founders, key executives — tagged per person. ``` Import these interviews into my veridive library, tagged '-leadership' and by executive name: — use veridive's library_import_youtube_video. ``` 3. **Run the behavioral questions** — Ask the questions a management presentation is designed to prevent. deep_context answers across the whole file, cited. ``` Across the '-leadership' videos: how do they talk about employee departures, internal failures and their customers? Quote the patterns, cite the moments. ``` 4. **Read the relationships** — The founder-to-founder read matters most in small targets — warmth, credit-sharing and consistency across tellings. ``` How do and describe each other and divide credit across these appearances? Anything that changed over time? Cited. ``` ## Problems it solves - Culture — the asset that decides post-acquisition retention — never appears in the data room. - Management presentations and curated references are theater by construction. - The behavioral tells are scattered across years of interviews nobody on the deal team has time to watch. - One-off impressions can't separate a bad interview day from a pattern; only the corpus view can. ## Why it's valuable - A behavioral read built from the leaders' own words across years — the layer the deal process hides. - Findings arrive cited: the IC hears 'here are four moments the CEO described departures this way,' not a hunch. - Deal-quiet: public record only, no outreach, no signal to the target. - The same file answers integration questions later — you'll know how this team communicates before day one. ## Example prompts ``` Import the leadership interviews, tagged per exec: — use veridive's library_import_youtube_video. ``` ``` Across '-leadership': how do they discuss departures, failures, customers? Patterns with citations. ``` ``` How do the founders describe each other across appearances? Changes over time? Cited. ``` ## FAQ **Isn't this reading tea leaves?** A single moment would be. The method is corpus-level: the same behavioral question across many appearances over time. Patterns that recur across contexts and years are how culture actually shows — and every finding stays cited so you can judge the context yourself. **What if the target's leadership has little public presence?** Common for smaller targets — then the sweep documents the thinness quickly, and the few appearances that exist get read carefully rather than not at all. Absence of record shifts weight to your in-person diligence; it doesn't waste it. **Which tools does this use?** library_import_youtube_video to build the tagged exhibit file, then deep_context to ask questions across it (scoped up to 20 items per question) with chunk-level citations. ## Related use cases - [Run founder due diligence from public appearances](https://veridive.com/mcp/use-cases/founder-due-diligence.md) — How does a founder talk about churn, failure and competitors across three years of interviews? Get the arc, cited to the minute. - [Catch M&A signals executives drop on podcasts](https://veridive.com/mcp/use-cases/podcast-ma-signals.md) — Long-form podcasts make executives chatty; sweep for the "we're very interested in that space" moments across your target list. - [Analyze how your spokesperson actually lands](https://veridive.com/mcp/use-cases/spokesperson-message-analysis.md) — Import their appearances, extract key quotes, and check delivered-message vs intended-message drift. --- Part of the veridive MCP use-case library: https://veridive.com/mcp/use-cases.md Server docs for agents: https://veridive.com/mcp.md