Quick links on this page

• Deep Firecrawl business application report
• Lead Claw Vision Layer pilot plan
• Extraction prompt + scoring schema
• CSV scorecard header

Guardrail: this is a research and internal execution page. No outreach, posting, paid API action, or customer contact happens from this page without Dad approval.

Firecrawl AI Clearly Explained — Deep Research + Business Application Report

Date: 2026-05-02

Video: https://www.youtube.com/watch?v=eH8JdttKIdA

Title: Firecrawl AI clearly explained (and how to make $$)

Channel: Greg Isenberg

Upload date: 2026-03-24

Runtime: 27m 28s

Local transcript: research/youtube/FIRECRAWLGREGISENBERGeH8JdttKIdA2026-05-02_transcript.txt

Metadata: research/youtube/videoeH8JdttKIdAmetadata.json

Cletus take — the real point for us

This is not just a “cool scraping tool” video. The big idea is:

AI without current, clean, structured data is smart but blind. The money is in giving AI reliable eyes, then packaging the output for a specific customer.

For us, that maps directly into Lead Claw, Revenue Leak Audit, SEO R.A.N.K., AnswerNeo, Hi-Tech Security, and Ohio LED Wall.

My recommendation: do not chase a generic SaaS yet. Start by adding Firecrawl-style website intelligence as a new layer on top of our existing Lead Claw workflow. Sell useful outputs first: scored prospects, audit snapshots, competitor gaps, missed-call/website leak reports, and niche monitoring. If a repeatable pattern sells, then we turn that into a small product/dashboard later.

Best internal name for the layer:

Lead Claw Vision Layer — Apify finds the businesses; Firecrawl reads their websites; Cletus scores the opportunity.

Source verification notes

Video claims checked

The video says Firecrawl turns websites into clean AI-ready data and supports scraping, crawling, mapping, search, agent-style data finding, and browser interaction. Firecrawl’s own docs/site broadly support that positioning:

• Firecrawl homepage says it helps AI systems search, scrape, and interact with the web, returning clean markdown, structured JSON, screenshots, and metadata.
• Docs list core capabilities: Search, Scrape, Interact, plus Agent, Browser Sandbox, Map, and Crawl.
• Docs describe /agent as early-access deep web data gathering where you describe what data you need; it searches/navigates and returns structured output.
• Docs describe /crawl as recursively discovering/scraping reachable subpages, with sitemap handling, JS rendering, path filters, depth limits, polling/webhooks, and credit billing.
• Pricing page currently shows: free one-time 500 credits; Hobby 3,000 credits/month at $16/mo billed yearly; Standard 100,000 credits/month at $83/mo billed yearly; Growth 500,000 credits/month at $333/mo billed yearly. Search = 1 credit/result, scrape = 1 credit/page, interact = 5 credits/action.

Important context / bias

The video is sponsored by Firecrawl. That does not make it useless — the strategic idea is strong — but we should treat the business claims as inspiration, not proof. We still need small controlled tests before depending on it for client work.

AI-agent hiring story

The video references Firecrawl hiring AI agents. TechCrunch reported Firecrawl posted “AI agents only” job ads, with a total $1M budget, including agent roles for content creation, customer support engineering, and junior developer work. The article also notes the catch: Firecrawl expects humans/companies behind many of these agents, and founder Caleb Peffer said today’s future is more like 10x engineers operating armies of agents, not magic human replacement.

That is useful for our positioning: we should sell AI-assisted systems supervised by humans, not fake “fully autonomous” hype.

Video breakdown by timestamp

0:00 — Intro / Thesis

Greg’s core line: Firecrawl is like giving AI eyes. AI models are useful, but they need current web context to do real work.

Practical translation for us:

• A local business’s website, Google profile, reviews, service pages, forms, hours, and competitor pages are all live signals.
• If we can extract those signals cleanly, we can produce valuable audits and sales intelligence fast.

2:14 — Why this matters now

He frames AI in eras:

1. Chatbot era — AI answers questions.
2. Copilot era — AI assists while humans drive.
3. Agent era — AI performs tasks, but needs web data.
4. Computer-use era — agents can browse, click, inspect, and act.

The money is not “AI can write text.” The money is AI can inspect reality and produce decisions/actions.

For us:

• Lead Claw should move from “list of businesses” to “business opportunity intelligence.”
• R.A.N.K. SEO should move from manual checklists to semi-automated site/competitor scans.
• Revenue Leak Audits can be pre-filled from public web data before a sales call.

7:40 — What Firecrawl is

Simple model:

Website / query -> Firecrawl -> clean markdown + JSON + screenshots + metadata -> LLM/report/dashboard

He lists six “superpowers”:

1. Scrape one page into clean markdown.
2. Crawl an entire site.
3. Map URLs on a domain.
4. Search and retrieve content.
5. Agent-style extraction from a plain-language request.
6. Browser-style interaction for clicks/forms/navigation.

For us, the highest-value parts are:

• Crawl: read a prospect’s whole site fast.
• Map: understand page structure and missing pages.
• Extract JSON: turn messy sites into scoreable fields.
• Screenshots: proof for reports.
• Agent: possible later for broader research tasks.

11:20 — How Firecrawl works / AWS analogy

Greg compares Firecrawl to AWS for web data: before AWS, builders managed servers; after AWS, one API call gave scalable infrastructure. His claim is Firecrawl does that for scraping/web data.

I’d soften that claim, but the business lesson is right:

The winners are not always the people who build infrastructure. They are the people who use new infrastructure to package a painful job better for a niche.

For us: we do not need to become “the Firecrawl company.” We use Firecrawl/Apify/OpenClaw to produce practical client-facing outputs for local businesses.

12:57 — Agent stack

Greg’s stack:

• Agent harness: Claude Code, Cursor, Codex, etc.
• Search layer: Perplexity, Exa, etc.
• Web data layer: Firecrawl.
• Ops brain: Obsidian, Notion, Apple Notes, etc.
• Outbound/audience layer: Instantly, Apollo, etc.

Our version:

Source acquisition: Apify / Google Maps / public web / directories

Web vision: Firecrawl / browser / direct fetch

Reasoning + production: OpenClaw + Cletus + Aims

Memory/ops brain: workspace docs + MEMORY + reports + Google Sheets

Offers: Revenue Leak Audit, SEO R.A.N.K., AnswerNeo, Hi-Tech Security, Ohio LED Wall

Approval/outbound: Dad-approved only

14:35 — Startup ideas

The video gives examples:

• Price monitoring.
• SEO gap finder.
• Job aggregation.
• AI research reports.
• Agent-in-a-box.
• Review intelligence.
• Lead-gen/enriched CSV batches.
• Real estate pricing data.
• SaaS competitor monitoring.
• Patent/legal filings.
• Influencer contact databases.
• Government contract alerts.

The recurring pattern:

Pick a niche -> collect hard-to-get data -> package the output -> sell the answer, not the tool -> automate repeat delivery

That is exactly what we should copy.

24:01 — Firecrawl hired an AI agent as employee

Greg uses this as a signal that companies will “hire” agent workflows for content, support, and junior developer tasks.

Our grounded version:

• Businesses will pay for done-for-you AI systems if the system saves labor, captures leads, or finds revenue.
• They do not really want “an AI agent.” They want calls answered, reviews monitored, quotes followed up, leads scored, reports generated, and tasks done.

26:24 — Final thoughts

The video ends on the “web data layer” opportunity. The big strategic takeaway is to build niche data products/services with high margins.

What this means for our businesses

1. Lead Claw gets a serious upgrade

Current Lead Claw direction: Apify pulls public business data, Cletus scores/ranks opportunities, exports CSV/call sheet/outreach notes, and Dad approves any outreach.

Add Firecrawl as the second layer:

Apify = find businesses and public listing data

Firecrawl = read each business website and competitor pages

Cletus/OpenClaw = score opportunity, summarize leaks, generate reports

Dad = approves outreach / offer direction

New Lead Claw fields Firecrawl can extract

For each prospect website:

• Does the phone number appear above the fold?
• Does the mobile page make calling easy?
• Is there an appointment/quote form?
• Does the form ask too much or too little?
• Are emergency/after-hours services mentioned?
• Are service pages separate or all lumped on one homepage?
• Are city/service-area pages present?
• Are testimonials/reviews/case studies present?
• Are project photos real or generic stock?
• Does the site mention AI chat, live chat, call answering, booking, CRM, or automation?
• Are title tags/meta descriptions clear?
• Is the page stack missing obvious money pages?
• What are the top 3 revenue leaks?
• What offer fits best: AnswerNeo, SEO R.A.N.K., website rebuild, camera/security, Wi-Fi/network cleanup?

This turns a raw lead list into a sales intelligence product.

2. Revenue Leak Audit becomes faster and sharper

Our existing Revenue Leak Audit already has a strong idea:

Stop selling AI. Sell the revenue leak.

Firecrawl can pre-fill much of the Snapshot before Dad or a client ever has a call.

Automated Revenue Leak Snapshot sections

For each business:

1. Contact path check — phone, form, CTA, booking link, after-hours clarity.
2. Missed-call risk — call-heavy service, emergency cues, no backup capture, no 24/7 answer path.
3. Website conversion leaks — weak CTA, hidden phone, slow quote process, confusing service copy.
4. SEO/page stack leaks — missing city/service pages, weak titles, thin content, no FAQs.
5. Trust leaks — lack of real photos, reviews, testimonials, case studies.
6. Best first fix — AnswerNeo, follow-up sheet, landing page, SEO page, form cleanup, etc.

Firecrawl gives us the site evidence; Cletus turns it into a useful report.

3. SEO R.A.N.K. gets an automation layer

Our R.A.N.K. formula is:

• Research the Money
• Assemble the Page Stack
• Network the Proof
• Keep Converting and Compounding

Firecrawl fits every step:

R — Research the Money

• Crawl competitor sites.
• Extract service pages.
• Identify which services/cities competitors emphasize.
• Compare page titles, CTAs, FAQs, and proof.

A — Assemble the Page Stack

• Map a client/prospect site.
• Detect missing pages: service, city, case study, FAQ, reviews, lead magnet.
• Generate a prioritized page list.

N — Network the Proof

• Check if project photos, testimonials, reviews, partner logos, schema, and social links are present.
• Build internal link suggestions.

K — Keep Converting and Compounding

• Re-crawl monthly.
• Track changes.
• Generate monthly “what improved / what is still leaking” reports.

This could become a monthly service:

Local Ranking Intelligence Report — monthly site + competitor crawl, action list, content/page opportunities, conversion leak scan.

4. AnswerNeo gets better prospect targeting

Firecrawl helps find prospects who *visibly need* call capture:

Good signals:

• Emergency service language.
• “Call now” everywhere but no after-hours system mentioned.
• Contact form only / no scheduling.
• High-ticket services.
• Many reviews, meaning real demand.
• Weak mobile CTA.
• Multiple locations or service areas.
• “24/7” claim with unclear routing.

For these prospects, our pitch is not “buy AI.” It is:

“We found a few places where calls and quote requests may leak. Want a quick snapshot?”

Firecrawl makes that snapshot evidence-based.

5. Hi-Tech Security opportunities

For Hi-Tech Security, we can use the same pattern for security-camera/access-control/network leads.

Possible Firecrawl-powered scans

Camera Blind Spot / Security Health Check prospecting

• Crawl local business sites for risk language: warehouse, yard, parking lot, after-hours, fleet, equipment, entry control, theft, vandalism.
• Combine with Google Maps business categories.
• Score likely camera/access-control fit.

Wi-Fi/network cleanup prospecting

• Find businesses with old websites, many locations, appointment/service model, and technology-heavy operations.
• Spot language around remote work, POS, online ordering, smart devices, guest Wi-Fi, camera systems.

Content intelligence

• Crawl competitor pages for topics they rank/sell around.
• Turn gaps into Hi-Tech service page/blog/reel ideas.

Proposal prep

• Before a site visit, crawl the prospect’s website and build a one-page “business context brief” so Dad walks in smarter.

6. Ohio LED Wall opportunities

Firecrawl can help us find LED wall rental/install leads.

Event lead radar

Scrape/crawl:

• Local event calendars.
• Festival pages.
• venue pages.
• school/church/community event pages.
• corporate event pages.
• wedding/expo/fair listings.

Extract:

• Event name.
• Date.
• Location.
• Expected attendance if public.
• Organizer/contact page.
• Sponsor/vendor info.
• Whether screens/stage/AV are mentioned.
• Fit score for LED wall rental.

Output:

• Weekly “Ohio LED Wall Event Lead Radar” with top opportunities.

Competitor monitor

Monitor competitor LED wall/rental/staging pages for:

• Pricing changes.
• Service area pages.
• New case studies.
• New gallery photos.
• Keywords they are pushing.

Use that to improve Ohio LED Wall pages, posts, and offers.

7. No Nursing Home / affiliate content

Firecrawl can help with content/affiliate research:

• Crawl product pages and comparison articles.
• Extract specs, common complaints, price ranges, safety features.
• Build updated buyer guides.
• Track broken/outdated affiliate product links.
• Identify content gaps against competing senior-safety sites.

Best use: content research and maintenance, not aggressive scraping.

Apify vs Firecrawl — how we should think about it

We already have Lead Claw started with Apify. This video does not mean “throw away Apify.” They fit different jobs.

Best stack:

Apify for “where are the businesses?”

Firecrawl for “what does their website reveal?”

OpenClaw/Cletus for “what does it mean and what should we sell?”

Best startup/service ideas for us from this video

Idea 1 — Lead Claw Website Intelligence Batch

What it is: Client gives a niche/city or list of businesses. We return enriched CSV + top 10 opportunity report.

Inputs: Google Maps/Apify list + Firecrawl website scans.

Output: scored leads, website leaks, offer fit, recommended first contact angle.

Buyer: local marketers, agencies, service businesses, our own sales pipeline.

Price: internal first; later $250–$1,500 per batch depending depth.

Why it fits: We already started Lead Claw.

Idea 2 — Revenue Leak Snapshot Factory

What it is: Firecrawl reads a prospect’s website; Cletus produces a 1-page snapshot showing missed-call/form/CTA/SEO trust leaks.

Output: one-page PDF/markdown snapshot.

Buyer: HVAC, plumbing, electrical, med spas, roofers, security companies.

Price: free/$49 lead magnet, then $500–$1,500 audit, then setup/monthly.

Why it fits: Directly supports AnswerNeo and AI Follow-Up & Call Capture.

Idea 3 — Local SEO Gap Finder by Niche

What it is: Crawl a business and top 3–5 competitors, then show missing pages/keywords/proof/CTAs.

Output: R.A.N.K. scorecard + page map + next 10 actions.

Buyer: local service businesses, agencies.

Price: $500 setup audit + $300–$1,500/mo monthly report/content support.

Why it fits: We just built the R.A.N.K. system.

Idea 4 — Ohio LED Wall Event Lead Radar

What it is: Weekly scan of public event/venue/festival/community pages for events likely to need LED wall rental.

Output: ranked event lead sheet + organizer/contact + reason fit.

Buyer: us first. Later could be sold to other event vendors.

Price: internal revenue tool first.

Why it fits: Direct path to rental/install opportunities.

Idea 5 — Local Competitor Watchtower

What it is: Monitor competitor pages for new services, pricing, case studies, SEO moves, and offers.

Output: monthly “what changed / what to copy or counter” report.

Buyer: our businesses first; later clients.

Price: $150–$500/mo add-on.

Why it fits: Easy recurring value.

Idea 6 — Review Intelligence for Local Services

What it is: Pull reviews/website data and summarize recurring complaints/opportunities.

Example: “Customers keep complaining about no callbacks, late arrivals, unclear estimates — here’s the offer/content/service fix.”

Buyer: local business owners, agencies, product/service teams.

Price: $99–$500/mo depending niche.

Why it fits: Helps AnswerNeo, service positioning, content, and ads.

Idea 7 — AI Agent Operator Services

What it is: Not selling “autonomous employees,” but supervised workflows that do a job weekly/monthly.

Possible agent roles:

• Content research agent.
• Prospect research agent.
• SEO gap agent.
• Competitor watch agent.
• Review intelligence agent.
• Support/call triage assistant.

Buyer: local businesses that cannot hire full-time marketing/ops help.

Price: $300–$2,000/mo depending task.

Why it fits: Aims Clawz / OpenClaw / AnswerNeo.

Recommended first pilot — no new SaaS, just prove the workflow

Pilot name

Lead Claw Vision Layer — 10 Prospect Test

Goal

Take 10 already-found local businesses and enrich them with website intelligence so we can produce better Revenue Leak / SEO / AnswerNeo angles.

Inputs

Use one existing Lead Claw category:

• HVAC contractors
• plumbers
• electricians
• roofers
• med spas

Process

1. Start with Apify/Google Maps lead list.
2. For each business with a website, use Firecrawl or a similar crawler to read up to 5–10 pages:

• homepage
• contact page
• service pages
• about page
• reviews/testimonials page if present

1. Extract structured fields.
2. Score opportunity.
3. Generate a 1-page internal snapshot.
4. Compare results to manual inspection.
5. Keep top 3 for Dad review.

Suggested JSON schema

{

  "business_name": "string",

  "website": "string",

  "phone_visible": true,

  "phone_above_fold_likely": true,

  "primary_cta": "string",

  "contact_form_present": true,

  "booking_link_present": false,

  "after_hours_language": "string",

  "emergency_service_language": "string",

  "services_detected": ["string"],

  "cities_or_service_areas_detected": ["string"],

  "missing_money_pages": ["string"],

  "trust_assets_present": ["reviews", "photos", "case_studies", "testimonials"],

  "seo_gaps": ["string"],

  "conversion_gaps": ["string"],

  "answerneo_fit_score_1_10": 0,

  "seo_rank_fit_score_1_10": 0,

  "website_rebuild_fit_score_1_10": 0,

  "hitech_security_fit_score_1_10": 0,

  "recommended_offer": "string",

  "one_sentence_pitch_angle": "string",

  "evidence_urls": ["string"]

}

Cost sanity

Firecrawl free plan currently lists 500 one-time credits. If scrape = 1 credit/page, a 10-prospect test at 10 pages each is ~100 credits before extra features. That is enough for a small proof without committing to a paid plan.

Our first Apify proof run cost was about $0.09 for 15 test leads, so the combined data cost can stay tiny if we cap pages and test carefully.

Success criteria

The pilot is worth continuing only if:

• It finds real, useful website/contact/SEO gaps.
• It saves us meaningful research time.
• The AI output is accurate after spot-checking.
• It improves our outreach/report angle.
• The cost per enriched lead stays low.

Offer packaging we can steal from this video

The video repeatedly says: sell the output, not the tool.

For us, that means:

Bad positioning:

“We use AI agents and Firecrawl to scrape your website.”

Good positioning:

“We found where your website may be leaking calls, quote requests, and Google rankings. Here’s the snapshot.”

Better positioning:

“We help local service businesses stop losing calls, quotes, and search traffic — with a quick audit, practical fixes, and optional AI follow-up.”

Possible offer names

• Website Revenue Leak Snapshot
• Local Search Gap Snapshot
• Lead Claw Website Intelligence Report
• AI Follow-Up & Call Capture Snapshot
• R.A.N.K. Local Growth Scan
• Camera Blind Spot + Website Trust Scan
• Event Lead Radar for Ohio LED Wall

Content ideas from this video

Short-form video hooks

1. “AI is smart, but it’s blind until it can read your website.”
2. “Your website might be leaking calls before you ever hear the phone ring.”
3. “We don’t sell AI. We find the business leak closest to revenue.”
4. “If Google, AI, and customers can’t understand your service pages, you’re invisible.”
5. “The next local-business advantage is clean data: reviews, services, pages, calls, and follow-up.”
6. “A website is not just a brochure anymore. It’s fuel for AI agents, search engines, and sales systems.”
7. “Most small businesses don’t need more software. They need someone to find the leaks and fix them.”
8. “The best AI business right now might be boring: read websites, find gaps, sell useful reports.”
9. “Local SEO isn’t magic. It’s page stack, proof, calls, and follow-up.”
10. “Lead lists are cheap. Scored opportunity intelligence is valuable.”

Blog/page topics

• “How to Tell If Your Website Is Leaking Calls”
• “What AI Sees When It Reads Your Local Business Website”
• “The Local SEO Page Stack Every Contractor Needs”
• “Why Missed Calls Are Usually a System Problem, Not an Employee Problem”
• “How We Score Local Businesses for Revenue Leaks”
• “Why Clean Business Data Is the New Marketing Advantage”

Guardrails before we use this seriously

1. Public/authorized data only. Do not scrape private/customer portals unless the client owns it and authorizes it.
2. Respect robots.txt and site terms where applicable. Firecrawl says its crawl endpoint respects robots.txt rules for FirecrawlAgent; still use common sense.
3. No spam. Lead Claw remains intelligence, not an excuse to blast people.
4. Dad approves outreach. Same rule as existing Lead Claw.
5. Human spot-check before client delivery. AI extraction can hallucinate or miss context.
6. Evidence URLs in every report. Every claim should point back to source pages/screenshots.
7. Cap cost per run. Limit crawl pages and credits during testing.
8. Do not overpromise autonomous agents. Sell supervised, practical workflows.

7-day action plan

Day 1 — Setup decision

Dad decides whether to create/use a Firecrawl account/API key for a small test. No public action needed.

Day 2 — Local setup

If approved:

• Save token privately in .secrets/firecrawl.env.
• Do not paste token into chat/docs/git.
• Install CLI/SDK if needed.
• Create a capped test script.

Day 3 — Test on our own sites first

Run against:

• 4-hitech.com
• hi-techsecuritysolutions.com
• ohioledwall.com
• aimsclawz.com
• answerneo.com

Output: R.A.N.K. style crawl notes and missing page/CTA issues.

Day 4 — Test on existing Lead Claw prospects

Use 5–10 prospects from the current Revenue Leak / AI assessment batches.

Day 5 — Generate top 3 snapshots

Create 3 sample internal snapshots. Do not send.

Day 6 — Compare manual vs automated accuracy

Spot-check every claim. Mark false positives/false negatives.

Day 7 — Decide whether to productize

Options:

1. Keep as internal research tool only.
2. Add it to Revenue Leak Audit workflow.
3. Package a paid “Website Revenue Leak Snapshot.”
4. Build a recurring monthly monitoring service.

30-day roadmap if pilot works

Week 1

• Finish 10-prospect pilot.
• Build repeatable extraction schema.
• Save prompt/templates.

Week 2

• Make a report template for Revenue Leak + R.A.N.K. combined snapshot.
• Build top-10 prospect review packet from enriched data.

Week 3

• Run one vertical batch: HVAC/plumbing/electrical OR med spas.
• Generate content around the most common leaks.

Week 4

• Create one sales page/landing section for the Snapshot offer.
• Decide on pricing:
• Free/$49 Snapshot
• $500–$1,500 Audit
• $1,500–$5,000 setup/fix package
• $300–$1,500/mo monitoring/AI Ops

My recommendation

This is a strong direction. The highest-value move is not “learn Firecrawl” as a hobby. The move is:

Add a web-data intelligence layer to Lead Claw so every lead becomes a scored opportunity with evidence, a clear offer angle, and a practical next step.

Priority order:

1. Lead Claw Vision Layer — enrich our existing leads with website intelligence.
2. Revenue Leak Snapshot Factory — create fast, evidence-based reports.
3. R.A.N.K. SEO Gap Finder — turn competitor/site crawl into SEO action plans.
4. Ohio LED Wall Event Lead Radar — find event opportunities weekly.
5. Competitor Watchtower — recurring monthly intelligence.

Do not build a dashboard first. Sell/report manually first. Once the workflow repeats and makes money, then automate deeper.

Files created from this research

• Transcript: research/youtube/FIRECRAWLGREGISENBERGeH8JdttKIdA2026-05-02_transcript.txt
• Metadata: research/youtube/videoeH8JdttKIdAmetadata.json
• This report: research/youtube/FIRECRAWLAIWEBDATALAYERGREGISENBERGeH8JdttKIdA2026-05-02.md

Source URLs

• Video: https://www.youtube.com/watch?v=eH8JdttKIdA
• Firecrawl homepage: https://www.firecrawl.dev/
• Firecrawl pricing: https://www.firecrawl.dev/pricing
• Firecrawl docs introduction: https://docs.firecrawl.dev/introduction
• Firecrawl Agent docs: https://docs.firecrawl.dev/features/agent.md
• Firecrawl Extract docs: https://docs.firecrawl.dev/features/extract.md
• Firecrawl Crawl docs: https://docs.firecrawl.dev/features/crawl.md
• Firecrawl MCP docs: https://docs.firecrawl.dev/mcp-server.md
• TechCrunch AI-agent hiring article: https://techcrunch.com/2025/05/17/y-combinator-startup-firecrawl-is-ready-to-pay-1m-to-hire-three-ai-agents-as-employees/
• Apify homepage: https://apify.com/
• Apify pricing: https://apify.com/pricing
• Apify platform docs: https://docs.apify.com/platform

Lead Claw Vision Layer — Firecrawl Pilot Plan

Created: 2026-05-02

Status: Internal proposal / no outreach / no paid API action unless Dad approves

Big idea

Add a website-reading layer to Lead Claw:

Apify finds businesses -> Firecrawl reads their websites -> Cletus scores leaks/opportunities -> Dad approves any outreach

This turns raw lead lists into scored opportunity intelligence.

Why this matters

Lead lists are cheap. A list that says *why this business is likely leaking calls, missing SEO pages, weak on CTAs, or ready for AnswerNeo/Hi-Tech/website help* is valuable.

First test

Use 10 local businesses from an existing Lead Claw/Revenue Leak batch.

Preferred niches:

1. HVAC / plumbing / electrical
2. med spas
3. roofers
4. auto repair / towing
5. security camera / access control prospects

Pages to crawl per prospect

Cap at 5–10 pages/business for the pilot:

• homepage
• contact page
• about page
• main service pages
• testimonials/reviews/project pages if present

Extraction fields

{

  "business_name": "string",

  "website": "string",

  "phone_visible": true,

  "phone_above_fold_likely": true,

  "primary_cta": "string",

  "contact_form_present": true,

  "booking_link_present": false,

  "after_hours_language": "string",

  "emergency_service_language": "string",

  "services_detected": ["string"],

  "cities_or_service_areas_detected": ["string"],

  "missing_money_pages": ["string"],

  "trust_assets_present": ["reviews", "photos", "case_studies", "testimonials"],

  "seo_gaps": ["string"],

  "conversion_gaps": ["string"],

  "answerneo_fit_score_1_10": 0,

  "seo_rank_fit_score_1_10": 0,

  "website_rebuild_fit_score_1_10": 0,

  "hitech_security_fit_score_1_10": 0,

  "recommended_offer": "string",

  "one_sentence_pitch_angle": "string",

  "evidence_urls": ["string"]

}

Scoring buckets

AnswerNeo / call capture fit

High if:

• emergency/after-hours language
• phone-heavy service
• high ticket value
• no clear backup capture
• weak contact form/follow-up path

SEO R.A.N.K. fit

High if:

• missing service/city pages
• weak title/meta/H1 signals
• few FAQs
• little proof/case-study content
• competitors have stronger page stacks

Website rebuild / conversion cleanup fit

High if:

• hidden phone/CTA
• confusing service copy
• old-looking site
• broken/weak forms
• poor mobile path likely

Hi-Tech Security fit

High if:

• business has facilities, parking lots, yards, inventory, equipment, or after-hours risk
• site mentions fleet, warehouse, access, employees, multiple locations
• category suggests cameras/access control/data/Wi-Fi needs

Output files for pilot

leads/lead_claw/vision_layer_runs/YYYY-MM-DD-niche/

  input-prospects.csv

  raw-firecrawl/

  enriched-prospects.json

  enriched-prospects.csv

  top-10-review.md

  snapshots/

Guardrails

• Public/authorized data only.
• Respect robots.txt/site terms where applicable.
• No outreach unless Dad approves exact prospect + copy + channel.
• Human spot-check before client-facing reports.
• Cap crawl pages and API cost during pilot.
• Every claim needs an evidence URL or screenshot.

Success criteria

Continue only if the pilot:

• saves research time
• finds real, useful leaks
• improves offer angles
• produces accurate outputs after spot-checking
• stays cheap per lead

Recommended first deliverable

A one-page internal snapshot per prospect:

1. Business + website
2. What they sell / service area
3. Contact/call path notes
4. Top 3 revenue leaks
5. Top 3 SEO/page gaps
6. Best-fit offer
7. One-sentence outreach angle
8. Evidence URLs

Related research

Full video/report: research/youtube/FIRECRAWLAIWEBDATALAYERGREGISENBERGeH8JdttKIdA2026-05-02.md

Lead Claw Vision Layer — Extraction Prompt + Scorecard Template

Created: 2026-05-02

Status: Internal reusable template

Use with: Firecrawl / browser scrape / manual site review / OpenClaw report drafting

Purpose

Turn a local business website into a scored opportunity profile for:

• AnswerNeo / AI Follow-Up & Call Capture
• Revenue Leak Snapshot
• SEO R.A.N.K. gap scan
• website/conversion cleanup
• Hi-Tech Security camera/access/Wi-Fi opportunities
• Ohio LED Wall/event opportunities where relevant

Do not send outreach automatically. This creates intelligence for Dad review.

Copy/paste extraction prompt

You are analyzing a local business website for practical revenue, SEO, call-capture, and service-fit opportunities.



Use only evidence from the provided website pages/content. Do not invent facts. If something is missing or unclear, say "unknown" or "not found".



Goal: produce a concise, evidence-based lead intelligence profile that helps decide whether this business is a fit for one of these offers:

1. AnswerNeo / AI Follow-Up & Call Capture

2. Revenue Leak Snapshot

3. SEO R.A.N.K. local ranking growth scan

4. Website/conversion cleanup

5. Hi-Tech Security cameras/access control/Wi-Fi/network cleanup

6. Ohio LED Wall/event/AV opportunity, if relevant



Return structured JSON plus a short human summary.



Look for:

- phone visibility and contact path

- above-the-fold CTA clarity

- quote/contact/booking forms

- emergency or after-hours service language

- service pages and city/service-area pages

- missing money pages

- reviews, testimonials, case studies, real project photos

- weak SEO/page structure signals

- trust gaps

- likely high-ticket services

- signs of facility/security/network needs

- best first offer angle

- evidence URLs for each important claim



Be practical and conservative. The output should help us decide whether to inspect manually and possibly ask Dad for outreach approval.

JSON schema

{

  "business_name": "string",

  "website": "string",

  "business_category": "string",

  "location_or_service_area": "string",

  "phone_visible": "yes/no/unknown",

  "phone_above_fold_likely": "yes/no/unknown",

  "primary_cta": "string or unknown",

  "contact_form_present": "yes/no/unknown",

  "booking_link_present": "yes/no/unknown",

  "after_hours_language": "string or not found",

  "emergency_service_language": "string or not found",

  "services_detected": ["string"],

  "cities_or_service_areas_detected": ["string"],

  "trust_assets_present": ["reviews", "testimonials", "case_studies", "real_photos", "certifications", "partner_logos"],

  "missing_money_pages": ["string"],

  "seo_gaps": ["string"],

  "conversion_gaps": ["string"],

  "security_or_network_fit_signals": ["string"],

  "event_or_led_wall_fit_signals": ["string"],

  "answerneo_fit_score_1_10": 0,

  "seo_rank_fit_score_1_10": 0,

  "website_cleanup_fit_score_1_10": 0,

  "hitech_security_fit_score_1_10": 0,

  "ohio_led_wall_fit_score_1_10": 0,

  "recommended_offer": "string",

  "top_3_leaks_or_gaps": ["string"],

  "one_sentence_pitch_angle": "string",

  "manual_review_needed": "yes/no",

  "risk_notes": ["string"],

  "evidence_urls": ["string"]

}

Scoring guide

AnswerNeo fit

Score high when:

• call-heavy or appointment-heavy service
• emergency/after-hours language
• high ticket jobs
• weak contact form or hidden phone
• no clear 24/7 answering / backup capture

SEO R.A.N.K. fit

Score high when:

• few/missing service pages
• weak city/service-area pages
• no FAQ/content depth
• few proof pages or case studies
• competitor pages likely stronger

Website cleanup fit

Score high when:

• CTA unclear
• phone hidden
• old/thin/confusing website
• forms hard to find
• poor proof/trust signals

Hi-Tech Security fit

Score high when:

• business has property, equipment, parking lots, warehouse, yard, vehicles, employees, multiple entrances, inventory, or after-hours risk
• category suggests cameras/access control/data cabling/Wi-Fi needs
• site mentions facility, fleet, operations, staff, deliveries, or commercial space

Ohio LED Wall fit

Score high when:

• event venue, school, church, festival, corporate event, entertainment, sports, wedding, stage, presentation, fundraiser, or outdoor crowd signals appear

Human summary format

Business: [name]

Website: [url]

Best-fit offer: [offer]

Fit score: AnswerNeo [x]/10 · SEO [x]/10 · Website [x]/10 · Hi-Tech [x]/10 · LED [x]/10



Why it may be a fit:

- [evidence-based point]

- [evidence-based point]

- [evidence-based point]



Top leaks/gaps:

1. [gap]

2. [gap]

3. [gap]



Suggested Dad-review angle:

"[one sentence]"



Manual review needed before outreach: yes

Evidence URLs:

- [url]

Guardrails

• Public/authorized pages only.
• Do not invent missing data.
• No outreach from this template alone.
• Every important claim needs evidence.
• Human spot-check before client-facing use.
• Dad approves prospect, copy, sender, and channel before external contact.

Lead Claw Vision Layer scorecard CSV header

business_name,website,business_category,location_or_service_area,phone_visible,phone_above_fold_likely,primary_cta,contact_form_present,booking_link_present,after_hours_language,emergency_service_language,services_detected,cities_or_service_areas_detected,trust_assets_present,missing_money_pages,seo_gaps,conversion_gaps,security_or_network_fit_signals,event_or_led_wall_fit_signals,answerneo_fit_score_1_10,seo_rank_fit_score_1_10,website_cleanup_fit_score_1_10,hitech_security_fit_score_1_10,ohio_led_wall_fit_score_1_10,recommended_offer,top_3_leaks_or_gaps,one_sentence_pitch_angle,manual_review_needed,risk_notes,evidence_urls,dad_approved_for_outreach,outreach_status