How to Do McKinsey-Level Analysis with AI (Without the $500K Engagement)

A Series A startup needed a competitive landscape analysis, market sizing, and go-to-market strategy for their board meeting. A top-tier consulting firm quoted $150,000 and an 8-week timeline.

The founder did it herself in a weekend using AI. The board called it "the most thorough competitive analysis we've seen from a startup at this stage."

Total cost: about $40 in API calls.

Here's the exact framework.

The Consulting Playbook, AI-Powered

Every major consulting engagement follows a structure: Define the question → Gather data → Analyze → Synthesize → Recommend → Present. AI accelerates every step.

Step 1: Market Sizing (TAM/SAM/SOM)

This is the bread and butter of consulting decks. With AI, it's a 30-minute exercise.

Perplexity: "What is the total addressable market for [industry] in 2026? Include recent analyst reports, industry data, and growth projections. Break down by segment and geography. Cite all sources."

Claude: Feed the Perplexity data and ask: "Build a bottom-up market sizing model for [our specific product/service]. Use these data points. Show the calculation for TAM, SAM, and SOM. Flag assumptions that are weak and suggest how to validate them."

Your judgment: Sanity-check the numbers against what you know from customer conversations. Adjust assumptions. The AI gives you the framework and the data; you provide the reality check.

Step 2: Competitive Landscape

Perplexity (multiple queries):

• "List all companies in [space]. Include founding year, total funding, headcount, and key differentiators."
• "What are the most common customer complaints about [competitor]? Check G2, Capterra, Reddit, and Twitter."
• "What has [competitor] launched in the past 6 months? Include pricing changes."

Claude: "Create a competitive positioning matrix using this data. Dimensions: [choose 2 strategic axes relevant to your market — e.g., 'ease of use vs. depth of functionality' or 'SMB focus vs. enterprise focus']. Place each competitor on the matrix. Identify the underserved quadrant where we have the strongest positioning."

This produces the 2x2 matrix that McKinsey charges $30K for. Yours is based on real customer data, not interviews with your VP of Sales.

Step 3: Customer Segmentation

Claude: "Based on this market data and our product capabilities, define 3-5 ideal customer segments. For each segment: describe the buyer persona, their primary pain point, the alternative they use today, our unique value proposition, estimated willingness to pay, and the best acquisition channel."

Power move: Upload your actual customer data (anonymized) — deal history, NPS scores, support tickets. "Analyze our existing customer base. Which segments have the highest LTV? Lowest churn? Fastest sales cycle? What patterns distinguish our best customers from our worst?"

Step 4: Strategic Recommendations

Here's where consulting firms justify their fees — the "so what" slide. AI does this well if you give it enough context.

Claude (with a role): "You are a senior partner at a top strategy firm. Based on this market sizing, competitive analysis, and customer segmentation, give me your top 3 strategic recommendations for the next 12 months. For each recommendation: the rationale, the expected impact, the risks, and the first 3 steps to execute. Be direct and opinionated — I'm paying you for judgment, not diplomacy."

Step 5: Board-Ready Presentation

Claude: "Turn this analysis into a 15-slide board presentation outline. Audience: Series A investors who care about market size, competitive moat, capital efficiency, and path to $10M ARR. Each slide: headline that makes one clear point, 3 supporting data points, and one visual suggestion."

Then use ChatGPT to generate any charts from your data, and drop everything into slides.

What This Looks Like in Practice

Monday evening (2 hours): Market sizing and competitive landscape research using Perplexity. Organize findings into a Claude Project.

Tuesday evening (2 hours): Deep analysis with Claude — competitive matrix, customer segmentation, strategic options.

Wednesday evening (2 hours): Synthesis, recommendations, and presentation outline. Final editing and slide creation.

Total: 6 hours over 3 evenings. A consulting firm charges $150K+ and takes 6-8 weeks for the same deliverable.

When You Still Need a Consultant

AI doesn't replace every consulting engagement:

• Industry relationships — if the real value is introductions, not analysis
• Political cover — sometimes the board wants "McKinsey said so" not "I used Claude"
• Implementation support — AI analyzes; consultants can also manage organizational change
• Proprietary data access — some firms have datasets you can't replicate

But for analytical work — market sizing, competitive analysis, strategic planning, pricing optimization, customer segmentation — AI produces comparable quality at 1/1000th the cost.

The Broader Implication

Every professional who can run this playbook becomes a one-person strategy team. Product managers can do their own market analysis. Founders can create investor-ready decks. Marketing leaders can build business cases for new initiatives.

The skill isn't the analysis itself — it's knowing which questions to ask and which assumptions to challenge. That's domain expertise. AI handles the research, data crunching, and presentation. You handle the insight.

Join AI Builder Club for strategy analysis templates and frameworks.