Using Claude and ChatGPT for Prospect Analysis

AI tools can dramatically compress the time needed to develop meaningful prospect insights. Used together, Claude and ChatGPT serve complementary roles in the ^[1] workflow: Claude for rapid company profiling from a URL, and ChatGPT for interpreting visual SEO data into agency-focused verdicts.

Claude: Company Profiling from a URL

The fastest way to understand a prospect's business model, market position, and recent news is to hand their URL directly to Claude.

Prompt pattern:

"Tell me about this company: [URL]"

Claude will read the site and return a structured summary covering:
- Core product or service offering
- Target market and distribution model
- Recent news or expansion activity (e.g., new retail partnerships)
- Key differentiators and brand positioning

Why this works: It replaces manually clicking through every page of a website. In under a minute you have enough context to speak intelligently about the prospect's business — including details (like a Whole Foods expansion or a B2B distribution model) that would be easy to miss in a quick browse.

When to use Gemini instead: Gemini tends to perform better when you need real-time web search results alongside the site content. For most company profiles, Claude is sufficient.

ChatGPT: Evaluating Ahrefs Screenshots

ChatGPT's vision capability makes it well-suited for interpreting full-page SEO dashboard screenshots and returning an agency-focused verdict.

Capture the Data

Use the Awesome Screenshot Chrome extension to take a full-page scrolling screenshot of the Ahrefs Site Explorer overview for the prospect's domain. This captures the entire dashboard — traffic trends, keyword counts, domain rating, backlink profile, and competitor data — in a single image.

Prompt for an Agency Verdict

Paste or upload the screenshot into ChatGPT and ask it to evaluate the prospect from a digital marketing agency's perspective. A well-configured ChatGPT (see ^[2]) will return analysis structured around:

• Overall SEO health — domain rating, authority headroom, backlink quality
• Traffic trends — growth, stagnation, or decline patterns and likely causes
• Keyword opportunity — positions 11–20 as quick wins, traffic capture rate vs. keywords ranked
• Paid search gaps — absence of PPC spend as a double opportunity
• Competitive context — how the prospect stacks up against named competitors
• Client fit verdict — a plain-language summary of whether and why to pursue the prospect

Example output (Twisted Alchemy): ChatGPT identified that the prospect ranked for 898 keywords but only captured traffic from 42% of them, averaged position 16 (bottom of page 2), had a keyword difficulty average of 5.96, and spent nothing on paid search. It concluded: "a very promising mid-market prospect — healthy but under-optimized authority, no paid search spend, and a clear need to invest to keep up with competitors."

Why Screenshots Over Raw Exports

CSV exports from Ahrefs can cause formatting issues when pasted into AI chat interfaces. PDF exports are more reliable, but a full-page screenshot is faster and preserves the visual context (charts, trend lines) that helps the model interpret trajectory, not just point-in-time numbers.

Configuring ChatGPT for Agency Context

ChatGPT's output quality improves significantly when it understands your role. Without context, it may frame analysis for a general audience rather than an agency evaluating a sales prospect.

Two approaches:

1. Custom Instructions (preferred): Go to Settings → Personalization → Custom Instructions and describe your role (e.g., "I run a digital marketing agency. I use this tool to evaluate potential clients for cold outreach. Always frame analysis from the perspective of an agency assessing whether and how to pitch a prospect."). After a few weeks of use, the model internalizes this context.

2. Session prompt: At the start of each conversation, include: "I am a marketing agency doing cold outreach. Evaluate this prospect from the perspective of an agency deciding whether to pitch them."

Workflow Integration

These AI steps fit between the ^[3] and ^[4] steps in the full research workflow:

1. Visit the prospect's website — note any obvious technical issues (e.g., missing naked domain redirect)
2. Claude: Paste the URL → get company profile
3. Ahrefs: Run Site Explorer → export keyword data
4. ChatGPT: Upload full-page Ahrefs screenshot → get agency verdict
5. SpyFu: Pull competitor traffic trends
6. Consolidate all findings into a prospect Google Doc (e.g., "Twisted Alchemy Insights - 1125")

The entire process — including both AI steps — takes 10–15 minutes per prospect.

Key Insight

The value of these tools is not automation — it's compression. You still make the judgment calls. Claude and ChatGPT eliminate the time spent reading every page of a website or manually interpreting a dashboard, so you arrive at the prospect conversation with specific, data-backed observations rather than generic talking points.

"You rank for almost 900 keywords, but only 42% of those keywords generate traffic. With strategic optimization of your position 11–20 keywords alone, you can deliver 5,500 additional visitors monthly, worth approximately $330,000 in annual revenue."

That kind of specificity — generated in minutes — is what makes insight-driven outreach work.

Related

• ^[1]
• ^[3]
• ^[4]
• ^[2]
• ^[5]
• ^[6]