CRIT Framework by Geoff Woods

CRIT stands for Context, Role, Interview, and Task—a structured AI prompting framework developed by leadership expert and bestselling author Geoff Woods. It transforms AI from a basic content generator into a strategic thought partner that helps legal professionals think deeper before typing.

Rather than asking an AI tool for an immediate answer, CRIT guides you through four intentional steps that surface better insights, catch edge cases, and produce more refined strategic decisions. Geoff Woods, founder of AI Leadership and host of the AI-Driven Leader podcast, built this framework on the principle that "the unfair advantage comes from how you think before you type."

Legal professionals across firm sizes—from solos to managing partners of 20+ attorney teams—are now using CRIT to analyze client matters, draft strategy, evaluate risk, and guide growth decisions.

The Interview component is described as the most valuable part of the CRIT Framework. Instead of asking AI for a direct answer, you instruct it to ask you clarifying questions first—typically one question at a time, up to three questions.

This shift fundamentally changes the quality of AI output. By asking AI to interview you before responding, you accomplish three things:

• Surface assumptions you didn't know you were making
• Clarify your actual goal (which is often different from your initial question)
• Enable AI to provide a more precise, contextual answer tailored to your specific situation

Legal professionals report that the Interview component improves output quality significantly, making it the lever most professionals underutilize.

Context (C): Provide detailed background about your situation, goals, constraints, and specific challenges. The richer the context, the more strategic the AI's response.

Role (R): Assign AI an expert identity. For example: "Act as a legal marketing consultant with 20 years of experience" or "You are a senior contract attorney at a Fortune 500 firm." Role-playing activates different reasoning patterns in the AI model.

Interview (I): Instruct AI to ask clarifying questions before answering: "Interview me. Ask one question at a time, up to three questions." Answer each question, then ask the AI to proceed with its response. This step is the single highest-value addition to basic AI usage.

Task (T): Define the specific, actionable deliverable you want. Instead of "Help me with pricing," ask "Create a tiered pricing model for three service packages, with columns for each tier, price range, and included services." Specificity drives output quality.

Law firms and solo practitioners are applying CRIT across a range of strategic challenges:

• Marketing Strategy: A managing director of an 8-attorney family law firm used CRIT to analyze Google Ads spend, diagnostic issues in lead cost, and campaign optimization—moving from reactive spending to data-driven allocation.
• Contract Review: A solo practitioner used CRIT to interview an AI about technology IP concerns in an NDA before drafting a response, ensuring no edge cases were missed.
• Growth Strategy: A managing partner of a 20-attorney personal injury firm used CRIT to evaluate expansion options, market readiness, and competitive positioning before committing capital.
• Client Matters: With proper safeguards and enterprise AI tools that don't train on input data, CRIT helps attorneys draft motions, analyze precedent, and structure arguments more strategically.

Implementation doesn't require technical expertise or extensive training. Start with these steps:

• Use Sticky Notes: Place reminders on your desk or monitor asking "How can AI help me do this?" and "Context, Role, Interview, Task." This keeps CRIT top-of-mind until it becomes habit.
• Start Low-Risk: Begin with internal strategic applications—analyzing your own workflows, planning, decision-making—before using CRIT on client-facing work. This builds confidence and surfaces what works for your practice.
• Create an AI Board of Advisors: Run the same question through AI with multiple expert roles (e.g., "You are a rainmaking partner" vs. "You are a risk-management specialist"). Compare the perspectives. This multiplies CRIT's value.
• Build Psychological Safety on Your Team: If you have associates or staff, frame CRIT experiments as learning, not productivity tests. Teams adopt AI tools faster when they feel safe experimenting.

Get a free AI visibility audit to see where your practice stands on AI readiness before rolling out CRIT across your team.

Most legal professionals who turn to ChatGPT or Claude ask a question and read the first response. This approach misses the full potential of modern AI models.

Basic usage: "Write a motion to compel discovery." → AI outputs a generic template.

CRIT usage: "Context: I'm the plaintiff's counsel in a products-liability case (case name, defendants, jurisdiction). Role: You're a discovery strategist with 15 years of experience. Interview: Ask me three questions about what we've already requested and what we're missing. Task: Write a motion to compel that addresses the gaps and cites controlling authority for our jurisdiction." → AI outputs a targeted, precedent-aware motion tailored to your case.

The difference is the structure. CRIT forces intentional thinking before asking AI to generate, which is why firms report time savings and higher output quality across decision-making.

Which AI platforms support CRIT?
CRIT works with ChatGPT, Claude, Google Gemini, and Perplexity. The framework is model-agnostic; it's the prompting structure that matters.

How long does it take to master CRIT?
The framework is learnable in minutes—you can grasp the four components in a single read. Mastery (where it becomes automatic) typically happens in 2–3 weeks of regular use.

Why does the Interview component improve output so much?
Interview forces AI to think through your situation before responding. It uncovers your unstated assumptions, clarifies your goal, and enables the AI to provide a response targeted to your actual need—not a generic answer to your surface question.

Is CRIT appropriate for client-facing work?
Yes, with proper safeguards. Use enterprise AI tools that explicitly don't train on your input data (e.g., ChatGPT Enterprise, Claude API with data privacy controls). Never use free-tier consumer AI for privileged communications or confidential client information.

Can solo practitioners use CRIT effectively?
Yes. Solo practitioners often see the fastest results because they make unilateral decisions and can implement new workflows immediately. CRIT fits naturally into a solo's day-to-day decision-making.

How much time can CRIT save me?
Legal professionals using AI report saving 1–5 hours weekly. Many report increased efficiency. The actual time savings depend on your current workflows and how extensively you adopt CRIT, but the gains are typically visible within the first month of consistent use.