Finding the right company is half the job.


Closing it is the other half — and it's not the same skill.

The RCG Deal Execution Framework is what happens as a sourced lead becomes a live conversation. Eighteen steps, four stages, from first call to closing — and every single one is marked for what it actually is: work that stays human because it takes relationship and judgment, or work AI now does faster and more thoroughly than a traditional deal team, with a person directing and approving everything it produces.

This is the Acquirers-stage version of the framework — eighteen steps, because acquisition carries the highest legal, financial, and negotiation complexity in the system. Earlier stages run the same Human/AI-Assisted/Both discipline, just at a fraction of the complexity — a first partnership or a sales pipeline doesn't need eighteen steps to execute well.

The Split

AI does the research.

A person does the deal.

Eight steps stay entirely human — fit assessment, NDA execution, engagement signature, LOI negotiation, closing. Relationship and judgment, start to finish, no shortcuts. Five are AI-heavy production work — financial spreading, the adjusted EBITDA build, data room setup, buyer research, CIM drafting — real analysis and drafting, always reviewed by a person before it's used. The remaining five are a direct blend, sequenced so each side does the part it's actually better at.

Four Stages

Lead → Live Conversation

Lead → Live Conversation. (2 Steps)

First outreach and fit assessment — a person makes the call, AI prepares the brief.

  1. Outreach & First Qualification Call — Human.

    A person makes the call, always; AI prepares the candidate brief beforehand.

  2. Deeper Fit Assessment — Human.

    Timeline, expectations, deal-breakers — reading a person, not just facts.

Conversation → NDA

Conversation → NDA. (2 Steps)

Securing the first real commitment and the first real financials.

  1. NDA / Confidentiality Agreement Execution — Human.

    The first real commitment a prospect makes, handled personally.

  2. Financial Document Collection — Both.

    A person asks; AI organizes what comes back and flags gaps immediately.

NDA → Engagement

NDA → Engagement. (5 Steps)

Financial modeling, a defensible value range, and a signed engagement — the work that turns interest into a real mandate.

  1. Financial Spreading & Baseline EBITDA Analysis — AI-Assisted.

    AI builds the model; a person reviews before anything is presented.

  2. Adjusted EBITDA & Addback Analysis — AI-Assisted.

    AI flags candidate addbacks; a person judges what's actually defensible.

  3. Value Range Development & Presentation — Human.

    AI builds the supporting comps; a person presents the range.

  4. Formal Engagement Terms Drafting — Both.

    AI drafts the terms; a person handles the real negotiation.

  5. Engagement Signature — Human.

    Final review, secured personally.

Engagement → Closing

Engagement → Closing. (9 Steps)

Buyer research, materials, negotiation, diligence, and close — the long stage, where the actual deal gets made.

  1. Data Room Setup — AI-Assisted.

    Organized and checked against a due-diligence checklist automatically.

  2. Buyer Universe Research & Target List — AI-Assisted.

    Built fresh for this engagement, reviewed and approved before any outreach.

  3. Teaser Development — Both.

    AI drafts it; a person checks positioning and confidentiality.

  4. CIM Development — AI-Assisted.

    AI drafts every section; a person ensures it's accurate and honest.

  5. Buyer Outreach & CA/NDA Management — Both.

    AI tracks the mechanics; a person qualifies real interest.

  6. Indications of Interest — Human.

    Every buyer conversation, every negotiation.

  7. LOI Negotiation — Human.

    Pure deal-making, start to finish.

  8. Due Diligence Coordination — Both.

    AI supports the mechanics; a person manages the real relationships.

  9. Closing — Human.

    Final negotiations and closing logistics, handled personally.

The Point

Eighteen steps. Not a claim about what AI can do — a map of exactly where it does the work, and exactly where a person still has to.