Hey everyone, currently Iam learning ai to scrape do deep research and find leads to offer my service which is local SEO. I have been trieng to build these system for weeks but my mind is messed up and everything is messed up. Could you guys tell me which tools you use for hunting emails or social profiles of the owner ?? Is It possible to create an automation, that scrapes google maps businneses, goes to their website find owners name, does deep research about the owner and also scrape the business social profiles to see If the owner if following his buinsess and then give It to me in a spreadcheat, is this possible? Also what tools do you guys use to deep research, I want to deep research to pick a qualified niche. Thats It, I hope someone can quide me with action steps, videos or reccomendations. Iam read to take action but I just dont know how.

I've just made the workspace I use to run local SEO clients free for everyone here. It's one Notion page you duplicate per client. Everything you need to take a business from kickoff to a paying monthly retainer lives in it: the prompts, the audit checklist, the fix list, the weekly rhythm, the citation directory and the client prospecting pack. Grab it here: https://nice-rooster-633.notion.site/Local-SEO-Client-Workspace-3c02a8320ad2814ba1efcc9d8194dd45 Duplicate it into your own Notion, rename it to the client's business name, and work top to bottom. SET THE STACK UP ONCE (DO THIS BEFORE ANYTHING ELSE) - Claude Code: the engine that runs every prompt in the doc - DataForSEO MCP: live SERPs, keywords, backlinks, on-page, Lighthouse - Windsor.ai MCP: the client's Search Console and GA4 in one connection, no API keys or scripts - Zernio MCP: Google Business Profile posts, metrics and review replies - Fathom: records the onboarding call so you never take notes - DataWise: point-and-click backlinks, competitors and keyword research, included with the AI Ranking membership Set it up once and it works for every client after that. The Prompts & Tools Library page in the doc has the setup links and a walkthrough video for connecting DataForSEO. HOW TO ACTUALLY USE THE DOC, IN ORDER 1. ONBOARDING The goal is one file, client.md, that every job you run for that client reads first. Run the "build the client file" prompt BEFORE you speak to them: Claude researches everything public, then turns every gap into a "> CONFIRM:" block. The call is you reading those blocks out loud. Record it with Fathom and paste the link into the Meetings log. Onboarding is done when this returns zero: grep -c "CONFIRM:" client.md

All six prompts from the video are attached as a zip. Plain markdown files, no code, no terminal. They run in the normal Claude app inside one Project per website. 🚨Prompt Are Attached - IF you are on mobile PLEASE scroll RIGHT🚨 What's inside 1. README — the setup in order, and the one rule 2. 00 Brief — run this first. It writes your site-brief.md, the file every other prompt reads 3. 01 Diagnose — finds what is actually costing you rankings, proves it with your own data, and gives you the fixes in order 4. 02 Check in — the numbers every two weeks and what moved. Schedule it 5. 03 Write — picks the next post from what Google already tried to send you, drafts it, then stops and asks about your experience 6. 04 Build — local businesses only. Where to get listed, free ones first, with your details block ready to paste 7. 05 Respond — drafts your review replies. Positive ones go out, negative ones wait for you Before you run anything Connect your data first. Windsor.ai for Search Console, Analytics and your Google Business Profile, then DataForSEO as a custom connector (the URL is in the README). The Brief prompt will not work without them, and Diagnose will just be guessing. Where things go The site-brief.md that Brief gives you gets uploaded to the Project's knowledge. The prompts get pasted as messages, or into the Project instructions if you want to just say "run diagnose". The one rule It proposes, you approve. Run every prompt manually the first week or two and correct it like a new hire. Once you like how it answers, tell Claude to save it as a skill for the Project and schedule it. Respond starts in dry run until you say "go live". Tools Used: - Windsor - DataForSEO Once the post is live, paste its URL into the description's free-prompts line if you want it to point straight at the post instead of the Starter Kit page.

Hi all, I've been struggling with finding local businesses to actually buy or lease my ranked sites, and I know a lot of people here have been too. I spent some time working through it last week, and here's a simple way I solved it for less than $10/m: 1. Scrape every business in your site's niche and city from Google Maps. I use a free Chrome extension (Map Lead Scraper, there are a dozen like it), takes 10 minutes and gives you name, phone, website and email for 100 to 300 businesses. 2. Sort the list by who's already paying for leads. Search your main keyword and note which businesses are running Google Ads but not ranking organically. Those are your buyers. They've already proven they'll pay for calls, they're just overpaying Google for them. 3. Download a free mail merge for Gmail (I use YAMM, free tier is lije 50 sends a day) and send a short plain-text email: "I own the site ranking #1 for X in Y. It gets Z calls a month. Want me to forward them to you for a week, free, so you can hear the quality yourself?" 4. This is the part that changed everything: put a Twilio number on the site (about $1/m) and actually forward the calls to your best prospect for 7 days. Then follow up with and ou're not pitching SEO anymore, you're taking away leads they've already tasted. Total cost: Twilio number $1.15/m, a Gmail address. Well under $10/m, and the first yes for me came witjin two weeks. If anything is unclear, let me know. Hope this helps you 🙏

Eight months ago Google didn't know this website existed (you can read the whole case study here) . One page, every service crammed into it, nav links that just scrolled you back to the same page. Today that business is the first name in the Google AI Overview for "restaurant accounting New York City" and the one ChatGPT recommends for the same search. I didn't build it. I told Steven what was wrong (Original Private Audit Video), gave him the plan, and he did the work. That plan is what's attached to this post. Five files, one per step, plus the launch checklist. Free, no email, nothing to buy. **On your phone? Scroll down, the zip is right there.** 5️⃣The five steps5️⃣ Every dead site is sick in a different way, so the steps are broad on purpose. Same as a doctor: the check-up is always the same, the medicine isn't. 1. Find what's actually wrong 👉🏼 Not just "no schema". The bigger question is whether this is a page problem or a structure problem. 2. Plan the fix 👉🏼 The page list and the checklist. Do this before you touch anything, because one fix contradicting another happens far more than people think. 3. Build it, then check nothing's missed 👉🏼 Fix by fix, ticking each one off. 4. Submit it to Google and Bing 👉🏼 Knock on the door. Sitemap in Search Console and Bing Webmaster Tools, or the single URL if you only fixed a page or two. 5. Track what happens. 👉🏼 Impressions first, then clicks, then conversions. If none of them move, go back to step 1. 🤐What's inside the zip🤐 1. README — which file does what, and how the checklist tabs map to the steps 2. 01 Find what's wrong — give it your URL (and any data you have). It hands you back a fix list and a verdict: pages or structure 3. 02 Plan the fix — takes that fix list and gives you the page list, one row per page, plus the checklist for each one 4. 03 Build and check — the build prompt, then the three checks on the live site: the quote paragraph, the schema, the proof