$300m of Meta Ads Hacks in Under 22 mins

Business outcomes / credibility claims

• Agency claim: $10.2B in sales generated, “mostly with Meta ads.”
• Ad spend scale claim: $300M+ spent across the agency; 1,000+ niches, 136 countries.
• Performance pattern cited repeatedly: tactics are said to 5x ROAS at times (exact results vary by example).

---

Hack #1: Use “phantom pages” (unbranded, persona-led ad destinations)

Problem with the standard approach

• “Professional” brand pages (logo, verified badge, polished grid) make ads look like ads in the feed → perceived as hard-sell / dead-on-arrival.

What to do instead

• Run Meta ads from unbranded “phantom pages” that look like people/personality/curator pages.
• Use founder-led pages: “People want to buy off people, not brands.”

Concrete examples

• Large spend brand (“Grounding Well”): >$100M/year sales mentioned.
  • Speaker highlights that persona page ads tend to be more aggressive/native.
• “Amber Hayes” example:
  • Non-branded ads appear editorial/advertorial (bridge → article → product),
  • with soft CTAs and minimal “ad-like” UI cues.

Actionable playbook

• Create a founder brand page (person-style presence).
• Take a winning ad from the brand page:
  • Do not change the ad
  • Republish the same ad creative from the founder/person page.
• Upgrade approach:
  • Run founder story ads from the founder page.

Framework / tactic label

• “Phantom page” (hack #1)

---

Hack #2: “Trojan story” — single-image + raw story ads (story > production)

Current behavior criticized

• Over-optimization around “creative” (hundreds of AI static ads, polished video production).

Speaker’s recommended creative

• A single, raw/static photo (phone photo / “camera roll” style).
• Pair with a mass-market story that creates curiosity/relatability.

Why it works (as described)

• The viewer is emotionally aligned with the creator’s “origin story.”
• Objections are reduced because the ad is framed as narrative, not a pitch.

Examples

• Speaker’s own: photo at desk; lead-in copy about making 150 cold calls/day, framed as a hero’s journey.
• “Amber Hayes” example:
  • Lead-in begins with a high-gossip hook (“I walked in on my wife cheating”),
  • then gradually connects to selling a grounding pillow.
• CTA pattern: “changed my life / find out more here” style →
  • typically advertorial/bridge style rather than direct hard selling.

Action steps

• “Write one story ad this week.”
• Use Reddit to find “mass market” stories/angles.
• Don’t start with “how do I sell”:
  • Start with the most compelling story action
  • Relate to the product at the end.

Framework / tactic label

• “Trojan story” (hack #2)

---

Hack #3: Optimize only the first ~11–14 words (“slippery leadin”)

Key concept

• Most people obsess over headline/body, but the real decision point is the top lead-in text the user sees to decide whether to click “See more.”

Example metric claim

• Changing lead-in from “450 leads per month…” to “I got 72 leads in 5 days” reportedly:
  • 5x ROAS

Rules for writing the lead-in

• Drop the viewer into the high point of the story (not the beginning).
• Make it conversational and dramatic/curiosity-based.
• Avoid “this is an ad” language and generic promotion.

Action steps

• In Ads Manager:
  • Take your top 3 best ads
  • Rewrite only the lead-in (keep the rest unless needed).

Framework / tactic label

• “Slippery leadin” (hack #3)

---

Hack #4: “Hyperdopamine” ads — copy the most-consumed content patterns on Meta

Input source

• Use Meta’s quarterly reports on:
  • most viewed content
  • most viewed pages
  • most followed pages

Creative model

• Use the same pattern Meta audiences already consume.
• Avoid generic clickbait by pairing intrigue with delivery.

Two ad-content buckets (as described)

• Breaking news
• Gossip

Ad framework given

• Big bold specific benefit (what the target wants)
• • “burning curiosity” (intrigue loop)

Checklist

• Does it interrupt their pattern?
• Does it open a curiosity loop?
• Does it include a hyperspecific big benefit?

Action steps

• Run your best ads through this checklist.
• Ensure the ad sells the click by delivering on what it teased.

Framework / tactic label

• “Hyperdopamine ad” (hack #4)

---

Hack #5: Use soft CTA “Learn more” (reduce friction)

Core idea

• Hard CTAs (“Shop now,” direct purchase) assume purchase intent exists.
• Conversion rate assumption cited:
  • 1–3% conversion
  • therefore most clicks are learning intent, not buying intent.

Recommended CTA

• Switch ads to “Learn more” (soft CTA).
• Logic: the landing page should do the hard selling; the ad reduces commitment.

Application

• In Ads Manager:
  • Change CTA to learn more
  • Let run 48 hours
  • Compare performance.

Framework / tactic label

• “Velvet CTA” (hack #5)

---

Hack #6: “ROAS barbell” — run only two creative extremes

Speaker’s positioning

• The middle (typical UGC/“normal” branded statics) is where most ads live → said to underperform.

Two creative buckets

1. Super ghetto statics: raw camera-roll images, minimal editing.
2. Go-ham productions: high-production skits, actors, exotic locations, heavy logistics.

What to avoid

• Middle-ground “normal UGC + standard branded layouts.”

Action

• Produce large volumes of raw statics (hundreds/thousands).
• Do periodic high-production “monster” shoots.

Framework / tactic label

• “ROAS barbell” (hack #6)

---

Hack #7: Use “zombie campaigns” to force Meta to spend on more ads

Problem

• Meta may refuse to spend on certain ads/batches even if they’re “good.”

Zombie campaign tactic

• Move loser ads (including high-conviction winners you expect to work) into their own CBO.
• Optionally create a dedicated budget per best subset.
• Goal: make Meta allocate spend; sometimes “winners emerge” from ads that were deprioritized.

Operational process

• Find a recent CBO campaign.
• Extract 80–90% of ads Meta wouldn’t spend on.
• Relaunch as:
  • a new zombie CBO, or
  • multiple zombie CBOs per ad set/batch.

Framework / tactic label

• “Zombie campaign” (hack #7, as described)

---

Hack #8: Never kill a dying ad—replace what fatigues first + use 7-day rolling decisions

Fatigue hierarchy (fastest → slowest)

• Image creative fatigues first
• Then headline
• Then body copy

Ad refresh playbook

• When a winner starts fatigue:
  • Keep body copy + headlines constant, but produce 15–100 new statics for the same message.
  • Or keep creative and change headlines (~50 new headlines).
  • Or keep creative/headline and change body copy (2–3 body variations).

Why not “pause daily”

• Don’t react to day-to-day ROAS volatility.
• Use 7-day rolling averages.
• For lead gen, ideally close the loop with lead metrics over ~1 week.

Three-step fatigue diagnosis

1. Avoid short-term reactions (no daily pausing/relaunching unless at extreme spend scale).
2. Check account-wide system health:
   • CPM, CTR
   • seasonal/economy events
   • major events hijacking attention like sports
3. Check frequency:
   • If frequency < 3 → unlikely fatigued
   • If above 3 → higher risk, but still not necessarily dead if the creative has “winning DNA.”

Action steps

• Create a 3-step sticky-note decision rule for media buyers.
• Don’t pause/kill until:
  • enough time has passed (at least 7 days of data), and
  • the frequency/health checks suggest true fatigue.

Framework / tactic label

• “Never kill a dying ad” (hack #8)

---

KPI / metrics explicitly mentioned

• Sales generated (agency claim): $10.2B
• Annual sales example brand: >$100M/year
• Ad spend (speaker claim): $300M+
• Conversion rate assumption: 1–3%
• ROAS improvements (examples):
  • “5x my ROAS” after lead-in change (72 leads in 5 days example)
• Ad testing timeboxes:
  • 48 hours after CTA change
• Decision window:
  • 7-day rolling averages
• Frequency thresholds:
  • <3 = not fatigued
  • >3 = higher fatigue risk

---

Presenters / sources

• Presenter / source of tactics: the video speaker (no name provided in subtitles) and their agency (referred to as King Kong / KingKong; also references “Subrey” as the speaker’s name in-part).
• Source referenced for platform data: Meta’s quarterly reports (most viewed pages/content, etc.).