I Asked 7 AI Models to Evaluate the Same Client. Here's Why Your Choice Matters
The business intelligence experiment that revealed how different AI models think about the same opportunity
Remember interviewing student workers? There was always one who over-researched, one who just wanted clear orders, and one who couldn't commit without getting feedback from everyone else. Turns out, AI models have the same personalities—and they make business decisions based on them.
Last week, I fed identical client prospect data into seven different AI models using the exact same prompt. What came back wasn't just different writing styles—it was seven completely different approaches to business intelligence. Some made executive decisions, others wrote investment-grade strategy docs, and one turned into a full-on research librarian—citations and all.
If you're using AI for business decisions (and you should be), understanding these personality differences could be the difference between getting actionable intelligence and... well, getting a very polite research paper.
The Experiment: One Prospect, Seven AI Personalities
I created a realistic prospect scenario based on common client profiles we see and asked each model to "transform this client research into a professional business intelligence brief for Kevin to evaluate this prospect."
Same data. Same prompt. Wildly different business intelligence.
I tested four Claude models, ChatGPT 4.0, Gemini 2.5, and Perplexity Sonar to get a comprehensive view of how different AI approaches business decision-making.
The Comparison: How Each AI "Thinks" About Business
Here's a quick cheat sheet to help you decide which AI to use depending on your business goal:
AI Model Personality Decision Style Length Key Strength Claude Sonnet 4 The Executive Makes the call for you ~500 words Strategic networking insights Claude Haiku 3.5 The Analyst Methodical, cautious ~350 words Systematic risk assessment Claude Sonnet 3.7 The Diplomat Balanced, measured ~400 words Relationship management focus Claude Opus The Investment Banker Strategic intelligence ~800 words Competitive landscape analysis ChatGPT 4.0 The Consultant Advisory tone ~450 words Practical implementation steps Gemini 2.5 The Senior Advisor Strategic counsel ~700 words Risk/opportunity matrices Perplexity Sonar The Researcher Data-driven ~800 words Market context with citations
The Personality Breakdown: Who Does What
Claude Sonnet 4: "The Executive"
Opening Line: "RECOMMENDATION: HIGH PRIORITY PROSPECT"
This AI doesn't mess around. It immediately spotted the networking goldmine (Sarah's Century 21 position = referral pipeline) and made a clear recommendation. Perfect for fast decision-making in high-pressure business development scenarios.
Best For: When you need someone to make the call and tell you why
Claude Haiku 3.5: "The Analyst"
Opening Line: "Prepared for: Kevin Harris, Harris Architect"
Pure methodical breakdown. Lists strengths, opportunities, and next steps without taking a position. It's the AI equivalent of "here's all the data, you decide."
Best For: When you want comprehensive analysis without bias
Claude Sonnet 3.7: "The Diplomat"
Opening Line: "The Thompsons represent a qualified prospect with a reasonable budget"
Balanced, relationship-focused, careful not to oversell or undersell. Noticed the "joint decision-making process likely" detail that others missed.
Best For: Client relationship management and nuanced situations
Claude Opus: "The Investment Banker"
Opening Line: "Sarah and Michael Thompson represent a well-qualified prospect for a $200,000 residential renovation"
Investment-grade analysis with strategic landscape mapping and actionable positioning advice. Actually calculated post-renovation property values and identified potential competitor firms.
Best For: When you need comprehensive strategic intelligence and competitive analysis
ChatGPT 4.0: "The Consultant"
Opening Line: "Business Intelligence Brief: Prospective Client Evaluation"
Structured like a professional consulting report with clear recommendations and implementation steps. Perfect for thorough client-facing documents that require both clarity and a structured, professional tone.
Best For: When you need a traditional business document format
Gemini 2.5: "The Senior Advisor"
Opening Line: "MEMORANDUM TO: Kevin Harris, Principal Architect"
Full memo format with sophisticated risk/opportunity matrices and strategic consultation advice. Anticipated client psychology and relationship dynamics better than others—like getting advice from a seasoned partner.
Best For: When you need senior-level strategic counsel with formal documentation
Perplexity Sonar: "The Researcher"
Opening Line: "Market Context: Current Market: Baton Rouge's housing market in 2025 is stabilizing..."
Went full research mode with real market data, citations, and comparative analysis. Created a market intelligence report instead of a prospect evaluation.
Best For: When you need comprehensive market research backing your decisions
The Most Revealing Differences
Strategic Thinking: Claude Opus wins for seeing the whole playing field
Only Sonnet 4 immediately flagged Sarah's Century 21 position as a referral pipeline opportunity. Claude Opus went further, mapping the entire competitive landscape and positioning strategy.
Winner: Claude Opus for comprehensive strategic intelligence Key Takeaway: For high-level business development, you'll want models that prioritize both networking opportunities and competitive analysis.
Risk Assessment: Gemini 2.5 and Opus offer structured frameworks
Sonnet 4: "LOW risk" with specific financial reasoning
Haiku 3.5: Listed systematic strengths/challenges
Gemini 2.5: Created formal risk/opportunity matrices Claude Opus: Identified specific red flags to monitor during the project
Winner: Gemini 2.5 and Claude Opus for sophisticated risk frameworks Key Takeaway: When you need thorough risk evaluation, these models provide the most structured assessment frameworks.
Market Intelligence: Perplexity and Opus back their opinions with data
Perplexity provided real market data with citations. Claude Opus calculated projected post-renovation values and analyzed neighborhood comps. Others worked with provided data only.
Winner: Tie between Perplexity Sonar and Claude Opus Key Takeaway: For data-backed decisions, these models transform prospect evaluation into comprehensive market analysis.
Decision-Making: Four strong models, pick based on your style
Sonnet 4: "HIGH PRIORITY PROSPECT" ✅
ChatGPT: Clear recommendations with action steps ✅
Claude Opus: "High-priority prospect" with strategic rationale ✅
Gemini 2.5: "Pursue this opportunity" with detailed positioning ✅
Sonnet 3.7: "Proceed with initial consultation" (safe middle ground)
Haiku 3.5: "Detailed consultation recommended" (punted)
Perplexity: Comprehensive analysis but no clear directive
Winner: Four-way tie for different decision-making styles Key Takeaway: When you need an AI to actually make the call, choose models that match your decision-making timeline and risk tolerance.
What This Means for Your Business
If you're using AI for business intelligence, your choice of model isn't just about writing quality—it's about decision-making frameworks.
Need quick strategic decisions? Sonnet 4
Want systematic analysis? Haiku 3.5
Managing sensitive relationships? Sonnet 3.7
Need investment-grade intelligence? Claude Opus
Creating client-facing reports? ChatGPT 4.0
Want senior-level strategic counsel? Gemini 2.5
Need market research backing? Perplexity Sonar
The Real Takeaway
The same data can lead to radically different insights—depending on which AI lens you choose. Some focused on financial risk, others on relationship dynamics, one on market positioning, another on networking opportunities, and others on competitive intelligence.
This isn't a bug—it's a feature. Just like hiring different types of analysts gives you different perspectives, using different AI models gives you multiple ways to think about the same business challenge.
Knowing which model fits your decision-making needs could be the key to unlocking more effective business strategies.
For Kevin Harris Architect, I now use Sonnet 4 for quick prospect evaluation, Claude Opus for strategic client intelligence, Perplexity for market research, and Sonnet 3.7 for client communication strategy. And yes, we now have seven AI subscriptions.
Have you tried using different AI models for your business? What differences have you noticed? Share your thoughts in the comments below, or let me know which AI personality fits your approach to business.
Trying to decide which AI tool to invest in? Test one scenario with several—then pick your team like a fantasy draft.
Curious how other AI models would handle this same challenge? If you have access to Microsoft Copilot, Meta's Llama, ChatGPT 3.5, or any other AI tools, run the same prompt and data through them - I'd love to see the results! Share in the comments or email me, and I might feature the most interesting responses in a follow-up post.
Want to see the full responses? Use the button above—it's fascinating to read how the same data gets interpreted through different AI frameworks.
*Elizabeth Harris handles business intelligence and AI integration at Kevin Harris Architect. When not conducting accidentally rigorous AI experiments, she's usually explaining to Kevin why we need seven different AI subscriptions. Subscribe for more insights on practical AI applications that actually help small businesses make better decisions.