Plus a 6-part prompt framework, PLR done properly, and why iPhone footage beat the £5,000 camera.

​If having trouble with any links, you can read the online version here​

Hey Guys,

One of today's videos contains a $10,000 lesson that took a guy about 4 weeks to learn. He was spending real money on Facebook ads, his pixel was firing, his leads were coming in, and every single one of them was training the algorithm to find him more time-wasters.

He fixed it with one piece of friction in the funnel. I'll show you where.

That's the theme running through all 10 videos today: the difference between doing the work and doing the work that counts.

TL;DR

Facebook now treats whatever hits your results column as your targeting, so let the wrong people through and you're paying to find more of them. Judge ads at the ad set level, because your worst-performing ad is often the one feeding all the others. Local businesses are leaking money in 5 obvious places and 1 white-labelled tool plugs all of them. Short-form has swung back to low-polish phone footage (640,000 views on an iPhone clip versus 68,000 on the professionally graded one). Nine principles turn a $4-an-hour writer into a $5,000-an-hour consultant. PLR works beautifully as raw material and terribly as a business model. Six prompts build a working AI task agent with auth and a chat bubble. AEO comes down to 5 moves and one 30-second robots.txt check. A tiny site hit 600 clicks on 4 repeatable content types. The RANKED framework fixes the prompts you're writing badly. And Perplexity's Personal Computer now touches your actual files across 400+ connected apps.

➡️ Your Complete Implementation Guide

Not got time to watch the videos? Grab the resources (Summary, Insights & Checklist) for each video from this Google doc

👉 ​Access Your Complete Implementation Guide Here​

1. Stop training Facebook to find you tyre-kickers

Dr. Matt Shiver has consulted 600+ service businesses and spends over $1,000 a day of his own money on ads. His first point is the one that cost him $10,000 to learn: whatever hits your results column is now your targeting.

What you'll discover:

• Why an application before the booking page (or a "which best describes you" question after the name and email) decides who trains your pixel, and how a 2-step form captures the lead first and qualifies second
• Why the ad soaking up 75% to 90% of your spend at a 1x return is holding up the 2x to 4x ads underneath it, and why you judge the ad set as a whole instead
• The attribution setting to switch off in 99% of campaigns, plus the compare-attribution column view that shows you how many "conversions" were really just people who scrolled past

🎥 Watch here​

2. The pitch that wins by giving the sale away first

Hamza Automates has built Hexona Systems into a 7-figure agency selling AI automation to plumbers, gyms and daycares. His pitch works because everyone else walks in quoting $5,000 and he walks in handing something over.

What you'll discover:

• The bucket model: 5 specific revenue leaks (missed calls, slow follow-up, no online booking, no reminder sequence, thin Google reviews) that every local business has and can see for themselves
• How to white-label 1 platform so it carries your logo and URL, then sell access at $100 to $300 a month, or take $2,000 to build it when the owner inevitably says "just do it for me"
• The after-hours test call that proves the problem before you ever pitch, and why a voice agent that books appointments takes about 5 minutes to build

🎥 Watch here​

3. Put the £5,000 camera down

James Smith has millions of short-form followers and just parted ways with his videographer. A meeting with Meta and 3 videos shot in Ireland convinced him the polished era is over.

What you'll discover:

• The real numbers from his feed: 68,000 views on the colour-graded edit, 640,000 on the clip he shot and trimmed on his phone, 136,000 on a petrol-station skit that took 7 minutes to film and 3 to cut
• The 5-step loop he now runs inside the Edits app (film, auto-captions, chop the fat, export, push everywhere) and why the whole thing fits in under 10 minutes
• The rep count nobody wants to hear: roughly 500 posts before you look like you know what you're doing, and he's at 5,000

🎥 Watch here​

4. Price versus cost, drawn on a whiteboard

Jason Fladlien went from writing articles at $4 an hour to charging $5,000 an hour. He breaks his persuasion system into 3 areas of 3 principles, and the shoe example alone is worth stealing.

What you'll discover:

• Emphasis: demonstrate quality without words, show the effort it takes to do it free (his webinar gave away 17 prompts and sold the other 26), and stage the pitch somewhere the value makes sense
• The $20 shoes worn 5 times at $4 a wear against the $200 shoes worn 600 times at 33 cents a wear, plus the question that reveals how much keeping the problem already costs
• The 5 development anchors (money, hours, effort, skill, risk) and the improve/approve/remove education loop that gets people saying yes before you ever pitch

🎥 Watch here​

5. PLR is the raw material, never the product

affiliatemarketingmc sold his first PLR product in 2000 and has watched the $497 resell-rights scene turn into what he calls a digital hot potato. His fix keeps the speed and drops the pyramid.

What you'll discover:

• The maths that kills most MRR offers: 88,000 members all reselling the same course at the same minimum price, so being first is the only edge that exists
• Why the constraint is the point, with a chili recipe book against 250,000 monthly searches and a press release guide against 28,000, both markets almost nobody is fighting over
• The 5-step build: pick the market, inject real value into the PLR, fan out 50 to 200 spoke pieces from the product's own problems and objections, capture emails, then run affiliate offers and a back end behind it

🎥 Watch here​

6. Six prompts, one working AI agent

Mikey No Code built a task app with private accounts and a conversational agent in 14 minutes, without writing code. If you've been circling Base44, this is the clearest run-through of the build order.

What you'll discover:

• The exact 6-prompt sequence: core task creation with 5 fields, filters plus a kanban layout, a dashboard with 4 summary cards and a progress bar, per-user authentication, a native chat bubble, then the upgrade to a context-aware agent
• Why each prompt adds 1 feature instead of 10, and how that keeps you inside the 25 free monthly credits while you learn what the platform does
• The difference between an assistant that follows commands and an agent that reads your task list, spots what's overdue, and confirms changes before making them

🎥 Watch here​

7. Getting cited when AI answers the question for you

Ahrefs compressed a 90-minute AEO course into 19 minutes. The headline stat: their AI search traffic was 0.5% of visits and drove 12.1% of signups.

What you'll discover:

• Query fan-out in practice, where 1 prompt becomes dozens of searches behind the scenes (ChatGPT's deep research ran 420 of them for a single question about a red phone case)
• The 4 writing rules that make your content chunkable: answer first, sections that stand alone, entity-rich phrasing, one idea per sentence, plus the finding that word count has near-zero correlation with citations
• The strongest single lever they measured (branded mentions on other people's pages, ahead of backlinks) and the 30-second robots.txt check that 5.9% of sites are quietly failing

🎥 Watch here​

8. Four gaps, four article types, 600 clicks

WordsAtScale runs a site in a niche Ahrefs rates as brutally hard, and it has walked through multiple core updates. The whole system is 4 repeatable moves he runs through Claude every morning.

What you'll discover:

• The 4 keyword gaps that sit under one roof: news gap (today's viral story), question gap (unanswered Reddit threads), product gap (new low-competition products), data gap (statistics pieces that pull links and citations)
• How a scheduled Claude task hands him 2 topics a day, 1 news piece plus 1 rotating gap type, each mapped to its own article format
• The plumbing behind it: programmatic indexing through 2 API-connected providers depending on whether he wants instant or drip-fed, and an automated link exchange that took DR to 23

🎥 Watch here​

9. The prompt that writes your prompts

Nico from AI Ranking has written thousands of SEO prompts. His conclusion: the model matters less than the structure and the real data you feed it.

What you'll discover:

• The RANKED framework in full: Role, Audience, Non-negotiables, Keyword data, Examples, Deliverables, and why the K carries most of the weight
• How to wire live SEO data into Claude through a custom connector so you stop pasting CSV exports into the chat
• The meta-prompt move: drop 1 file in, ask for a keyword research prompt or a competitor analysis prompt, then run the output in a fresh window (the same prompt works in ChatGPT too)

🎥 Watch here​

10. Perplexity now works on your actual files

Julian Goldie walks through Personal Computer, which landed on Windows on July 28th after starting on Mac. It reads your approved folders, drives your apps, and connects to 400+ tools.

What you'll discover:

• What it actually does: takes an objective, splits it across 15+ models as sub-agents, remembers past tasks through a memory layer, and can drive the Comet browser to fill forms and book appointments
• The 2 prompts he gives you, both built the same way (name every section, name the format, name where it saves) because vague input produces vague output
• The 7 beginner guardrails, starting with approving 1 project folder instead of your whole drive and reading every permission alert instead of clicking through on autopilot

🎥 Watch here​

Talk soon,

- Tony

P.S. Video 7's freshness data is the nudge to act on today: 76% of ChatGPT's top-cited pages were refreshed within the last 30 days. Whatever you published 6 months ago is already slipping. Pick your best 3 posts and update them before the weekend.