Insight-Driven Research Workflow

Overview

Rather than sending generic automated outreach, this workflow produces a prospect-specific research document in under 15 minutes. The output is a Google Doc containing concrete SEO findings, competitor context, and AI-synthesized insights that can be used across all outreach touchpoints — cold emails, LinkedIn messages, voicemail follow-ups, and discovery calls.

The core principle: lead with value before asking for anything. When a prospect sees that you've analyzed their site and can name specific numbers (keyword positions, traffic gaps, revenue opportunities), they recognize competence and are far more likely to engage.

"If you lead by saying, hey, here's an insight that I found and I thought you might find it interesting — now the guy's like, huh, you gave me something." — Mark Hope

When to Use This

Use this workflow before any first contact with a qualified prospect. The resulting Google Doc becomes the single reference asset for all outreach to that company — emails, calls, and the eventual discovery call with a senior team member.

Step-by-Step Process

1. Visit the Prospect's Website

Before opening any tool, spend 2–3 minutes on the site itself.

• Does it load quickly and look professional?
• Check whether the naked domain (e.g., company.com) redirects to www.company.com. If it doesn't, that's an immediate, concrete finding to lead with.
• Note the product/service mix and any obvious content gaps.

2. Get a Business Summary via Claude

Paste the company URL into Claude and ask it to summarize the business:

"Tell me about this company: [URL]"

Claude will read the site and return a concise overview — business model, products, distribution channels, recent news, and market positioning. This saves you from manually reviewing every page and often surfaces non-obvious details (e.g., a recent Whole Foods expansion, a B2B distribution model).

Use Gemini as an alternative if you need more current web-sourced information.

3. Run Ahrefs Site Explorer

Enter the prospect's domain in Ahrefs → Site Explorer.

Key metrics to capture:

Export the keyword data as a Google Sheet or PDF. PDFs tend to parse better when dropped into AI tools.

Take a full-page screenshot of the Ahrefs overview using the Awesome Screenshot Chrome extension. This screenshot is used in Step 5.

Constructing the Opportunity Estimate

If you find keywords clustered in positions 11–20 with low difficulty:

1. Estimate additional monthly visitors if those keywords reach page 1
2. Apply a 2% conversion rate and the prospect's estimated average order value
3. Annualize → this becomes your headline number (e.g., "$330k annual revenue opportunity")

4. Run SpyFu Competitor Analysis

Search the prospect's domain in SpyFu to identify their top organic competitors.

• Note which competitors are growing vs. declining
• Look for market-wide trends (e.g., a sector-wide traffic drop since mid-2024 likely reflects AI search influence)
• Identify the top 3–5 competitors by name — you'll reference these in outreach

Strategic framing by scenario:

• Prospect is growing slightly, competitors declining → "There's market share to capture as your competitors lose ground"
• Prospect is declining alongside competitors → "We know from others in your space that traffic has been stagnating — here's how we'd help you recover"

Export the competitor list to Google Sheets for the research doc.

5. AI Synthesis via ChatGPT (Screenshot Analysis)

Drop the full-page Ahrefs screenshot into ChatGPT and ask for an agency-focused verdict.

Before doing this, configure ChatGPT's Custom Instructions (Settings → Personalization) to establish your agency context. This ensures responses are framed for a marketing agency doing prospect evaluation — not for a generic user. Alternatively, prepend each session with:

"I am a marketing agency doing cold outreach to potential clients. Analyze this from the perspective of whether this company is a good prospect and what I could help them with."

ChatGPT will return a structured assessment covering:
- Domain authority and headroom
- Traffic trends and likely causes
- Keyword intent mix (informational vs. commercial/transactional)
- Competitive positioning
- Paid search gaps
- A "client fit verdict" summarizing the opportunity

Documenting the Research

Create a Google Doc per prospect and paste everything in — raw exports, AI outputs, competitor lists, and your own notes. Don't over-format it; just get the content in.

Naming convention: [Company Name] Insights - [MMDD]
Example: Twisted Alchemy Insights - 1125

The date suffix prevents confusion if you revisit or refresh the research later.

Attach the Google Doc to the prospect's HubSpot record so it's accessible during calls and email drafting.

Using the Research in Outreach

Don't give everything away at once. Drip insights across a multi-touch sequence:

• Email 1: Lead with one specific finding (e.g., the naked domain redirect issue, or a single keyword gap)
• Email 2 (4–5 days later): Share a second insight (e.g., the $X revenue opportunity from position 11–20 keywords)
• Email 3: Reference a competitor by name and what's happening to their traffic

For the intro/follow-up template, include a placeholder for one insight and a 15-minute meeting ask. A guarantee framing works well:

"Give me 15 minutes — I guarantee you won't be sorry."

For LinkedIn outreach, the same principle applies: name the company, share one observation, offer to share more.

Handoff to Senior Discovery Call

Once a prospect expresses interest:

1. Jacob qualifies on a brief call — confirm they're willing to engage
2. Jacob books directly onto Mark's calendar using the calendar link (don't send the link to the prospect and wait — look at available slots together and make the appointment on the spot)
3. Mark leads the discovery call, using the research doc to present findings and move the deal forward

Tools & Access

Time Estimate

A complete research pass takes 10–15 minutes per prospect. The resulting Google Doc is reusable across all outreach to that company and eliminates the need to re-run research before calls.

Related

• ^[1]
• ^[2]
• ^[3]