ChatGPT Has Officially Changed Facebook Ads Forever! (Tutorial)

Core Idea

Professor Charley T argues that scaling Meta ads is less about launching more ads and more about building a growing pool of people who already know the business. His proposed system uses three prompts to attract relevant attention, identify which content earns followers efficiently, and sell to warm audiences through the existing sales campaign.

The approach is presented as a way to reduce repetitive Ads Manager work—not as a guarantee of lower acquisition costs or better results.

The Operating System

Campaign 1: Engagement

• Promote organic posts to people who are not already followers, using a small, controlled budget.
• Judge posts by whether they earn followers and at what cost per follow, rather than by cheap likes.
• Give posts a consistent spending or impression cap so each gets a comparable test.

Campaign 2: Sales

• Keep the existing broad, CBO sales campaign.
• Add a broad retargeting ad set inside it, aimed at people who have recently visited the site or engaged with the brand’s Facebook or Instagram content.
• The presenter argues that Meta’s campaign-level objectives mean the engagement campaign should remain separate, while retargeting belongs inside the sales campaign—not in an additional retargeting campaign.

The Three Prompts

1. Fill the audience: Promote selected organic posts to attract relevant followers.
2. Lower the cost of audience growth: Assess whether a post merits another run.
3. Convert warm audiences: Build a broad-retargeting ad set inside the existing sales campaign.

The prompts are designed to inspect the account, reuse what already exists, and ask for approval before publishing. The presenter demonstrates them through ChatGPT’s browser control of Ads Manager, with the user retaining control over credentials and live changes.

Playbooks and Decision Rules

• Audience windows: Use 30 days as a baseline. Shorter windows, such as 7 or 14 days, are the efficiency end, while longer windows, such as 90 or 180 days, increase reach. Widen the window as the pool grows or frequency rises.
• Content testing: Promote posts consistently and apply the same cap to make results comparable. Likes alone do not measure success.
• Post-recycling gates: The demonstrated skill considers a post for another run only if its latest run:
  • Was switched off by the existing automated rule.
  • Has no active copy still running.
  • Generated at least two follows and beat the account’s cost-per-follow baseline.
  • Reaches the tool’s stated 90% confidence threshold.

In the example, no posts qualified, so nothing was duplicated or published. The presenter stresses that a past win does not mean a post will keep winning.

• Retargeting setup: Check the actual audience, pixel, conversion event, location, and age settings. Verify that audiences are being used as targeting—not merely as Meta “suggestions” that can expand beyond the warm pool.
• Performance evaluation: Judge the warm ad set by CPA and sales, not CPM alone. Warm impressions may cost more, but the presenter says the audience may convert more efficiently because people already recognize the brand.
• Automation safeguards: Review audience membership, exclusions, rules, ad status, and drafts. The demo catches issues such as an existing cap that would override a requested cap, a mismatched Instagram audience, and Meta adding shop visitors to a website audience.
• Starting from zero: Focus first on content worth promoting, then build the audience pool, then sell to it.

Metrics and Examples Cited

• The presenter illustrates cold-acquisition economics with a $50 cost per purchase and $20 CPM: 2,500 impressions per purchase.
• Demonstrated budgets: $50/day for engagement alongside $300/day for the sales campaign.
• The account reportedly had 15 warm audiences across 7-, 14-, 30-, 90-, and 180-day windows: 11 built and 4 reused.
• A warm audience estimate in the demo was 47,000–55,200 people, with one audience still populating. The presenter notes that Meta may label even a roughly 50,000-person audience as small.
• For the presenter’s follower campaign, the cited total was $986.28 spent for 42 follows, or about $23.48 per follow. He recommends using total spend divided by total follows—including poor-performing ads—rather than cherry-picking winners.
• An existing automated rule turned ads off after 250,000 impressions, checking approximately every 30–60 minutes. The tool initially offered a one-million-impression option, but the presenter chose to use the existing rule instead.
• The demonstrated live retargeting ad set had spent $6,698 in the previous 30 days. The presenter cited four live ads, including one at $3,375, one at $1,216, and two customer-review ads at about $390 each.
• The presenter claims his business, Disruptor Academy, grew 10× in just over a year and was returning about 6:1 on ad spend. These are his reported results, not a forecast for other advertisers.

Practical Cautions

• Treat results as account-specific. The presenter repeatedly says that costs and outcomes vary, and that the sales campaign—not follower growth by itself—shows whether the audience is valuable.
• Keep the final decision and publishing switch under human control. The demonstration leaves campaigns off and in drafts until approved.
• Check account defaults and automation rules rather than assuming they match the intended setup.
• Do not judge a warm campaign by CPM alone; assess whether it improves CPA and sales. If it does not, the presenter recommends reviewing whether the content is attracting the right people before abandoning the strategy.
• The video’s headline says ChatGPT “changed Facebook ads forever,” but the practical claim is narrower: browser automation can perform setup work, while the advertiser still needs to understand the strategy and make decisions.

Presenter and Sources

Presenter: Professor Charley T

Creators referenced in the video: Sam Pillier, Chase Chappelle, Ben Heath, and Jeano Cataccliff (name spelling as rendered in the subtitles).