← The Research Station ~ dish 04 of 10 · the research station

Vendor and Partner Due Diligence

Background, financial signals, customer reviews, founder track record. Assembled into a one-pager. The station researches. The Chef reads and decides. The leverage is guardrails on sources. Public. Sourced. Dated. No "the AI said so." Without that guardrail, rumor becomes fact.

The job

Before the business signs a contract with a new vendor or partner, someone needs to do due diligence. Is the company stable. Do customers trust them. Is the leadership credible. Right now that work is manual, inconsistent, and sometimes missed entirely. A quick call. A few Google searches. Maybe a review check. Then you sign and hope.

When the station runs this dish well, 48 hours before the contract is signed, the Chef gets a one-page due diligence summary. Company background. Financial signals (if public). Customer reviews (sourced). Founder or CEO background. Any red flags worth investigating further. The Chef reads it and makes an informed decision.

The difference between research that informs a decision and research that confabulates is knowing your sources. A rumor from a Slack group is not due diligence. A news article about the founder is. A customer review from a verified source is. A claim the station made up is not. Define your sources before the station ever runs.

The recipe

All seven ingredients still apply. The leverage on this dish is Guardrails (Ingredient #3). Stick to public, verifiable sources. Everything sourced. Everything dated. No “the AI said so.” Without that guardrail, the due diligence is theater.

Context matters. The station knows what information about a vendor is relevant to your business. Examples matter. Show the station a due diligence report you’d actually use to make a decision. What sources. What level of detail. Guardrails matter. Define what counts as a credible source (public filings, news, verified reviews) versus a weak source (rumor, unverified claims, third-party chatter).

How to build it

  1. Define the categories of due diligence relevant to your business. For a SaaS company buying a new tool, maybe it’s company stability, customer base, support reputation, and founder credibility. For a services firm hiring a partner, maybe it’s financial stability, customer satisfaction, and team depth.
  2. Identify the sources you trust for each category. Company background: website, LinkedIn, news. Financials: Crunchbase (if startup), SEC filings (if public). Reviews: G2, Capterra, or industry-specific sites. Founder: LinkedIn, news, previous roles. Be specific about which review site you trust.
  3. Define what “red flag” means. For a vendor, maybe it’s negative reviews from multiple customers, founder changes, or obvious financial distress signals. For a partner, maybe it’s high turnover, customer churn signals, or legal issues. Write these down.
  4. Define your due diligence format. Company name. Founded. Leadership (name, background, one line). Customer reviews (scored, sourced site). Financial signals (if available, dated). Red flags (if any). That’s it.
  5. Pull two due diligence reports on vendors you’ve used. This is your standard.
  6. Test on a mock run. Give the station three companies you’re considering a partnership with. It generates due diligence reports. You read them. Do they surface real signals or confabulation. If you see claims without sources, the guardrails aren’t tight enough. Adjust.
  7. Go live. Every report should include the source for every claim. No claims without backing. If a claim can’t be sourced, it doesn’t go in the report.

What breaks it

  • Sourcing includes rumor. The station includes a claim about a vendor’s financial health based on a GitHub comment or a Slack-like forum. Rumor becomes fact because it’s in the report. Stick to public, verifiable sources only. SEC filings. News articles. Verified customer reviews on trusted sites.
  • Customer reviews are unverified. The station pulls five-star reviews from somewhere without verifying that the reviewers are real customers. The report looks positive. You sign the deal. Three months in, you realize the reviews were fake. Always verify the review source. Is it G2, which has anti-fraud measures. Is it Capterra. Is it a random site with no verification.
  • Financial signals are stale. The report cites funding from three years ago and doesn’t mention recent developments. You assume the company is stable. You sign the deal. Two months later, the vendor closes. Get fresh signals. Recent news. Recent reviews. Recent hiring or customer signals.
  • Red flags are missed. There’s a news article about the vendor’s CEO leaving. The station doesn’t mention it. You sign the deal. A month later, the service degrades because of the transition. Be thorough. Check news. Check LinkedIn for team turnover. Check customer reviews for recent complaints.
  • Report includes sourcing but the sources are wrong. The station cites a source that doesn’t actually contain the claim. You don’t have time to fact-check every claim so you trust it. Then you discover the citation is wrong. Spot-check the sources. If 20 percent of claims don’t actually appear in the source cited, the recipe is hallucinating. Tighten.

When it’s working

At week four, when the Chef is considering a new vendor or partner, the Chef asks the station for due diligence. The station returns a one-page report in 48 hours. Every claim is sourced. Every source is verifiable. Red flags, if any, are clearly called out. The Chef reads it and makes a decision. The decision is more confident because the due diligence was thorough.

The signal that the recipe is sharp: after signing a partnership based on a due diligence report, the Chef doesn’t discover negative information that should have been in the report.

Monday Move

Define the categories of due diligence relevant to your business. Identify the sources you trust for each. Define what “red flag” looks like. Create a due diligence template. Pull one due diligence report on a vendor or partner you know well. The station is running on Monday.

Dish 4 of 10 on the Research Station. Build-note leverage: Guardrails (Ingredient #3).