Projects · Closed Podcast · Part 2

THE PACKAGING PLAYBOOK.

what the titles and thumbnails are doing psychologically, and the format the next episodes get written against.

Part 1 showed the reach on this channel is bought. This page reads the one thing the budget cannot buy: whether the title and thumbnail make someone choose the episode, and whether the episode keeps the promise. Two signals, both from YouTube Studio. Click-through rate is views from YouTube's own impressions, which ad views do not touch, so it is the clean read of the packaging. Average percentage viewed is whether the promise was kept. 34 promoted episodes, so read the buckets, not the rows. About 99 percent of the channel's traffic is YouTube advertising in both windows, so every bucket is confounded by promotion to some degree; anything that clusters in Season 3 (heavy spend, short thumbnails, pattern-interrupt titles) is still marked test, not use.

Title triggers

Channel medians: 2.5 percent click-through, 8.25 percent viewed. A click-through gap under about 0.3 points is noise at this sample size. Three title triggers are settled, not open: question marks, second person (you/your), and titles of 55 characters or fewer. The 2026-09-02 retitle removed all three from the catalog, so they sit at n=0 and are dropped from the table below rather than shown with numbers that no longer exist. Open loop is down to a single title (n=1) with no usable with/without split at that sample size; it stays a test, not a conclusion. The insider-word thumbnail bucket (n=2) still has a click-through read (2.25 vs 2.45 percent, avoid) but no reliable retention split to report at n=2. Two of the newest episodes, Brenda Powell and Sergio Diaz, are promoted with lifetime CTR but have no thumbnail tag yet, so they are excluded from the thumbnail-trigger table.

Trigger	n	Click-through, with vs without	Percent viewed, with vs without	Verdict
Guest's name in the titleNow in 33 of 34 titles, so it is the format, not a testable lever. The lone without is a pipe-delimiter pull artifact, not a real omission. Retiring the old biggest-lift claim.	33	2.5% vs 2.8%	8.4% vs 4.6%	Neutral
Third-person story (he, she, this agent)The live title lever: a 0.5-point click-through lift, above the noise floor.	15	2.8% vs 2.3%	8.1% vs 8.4%	Use
Two-part title: story, then payoffClick is mildly positive, but retention flipped negative against the old read. It is now the house style (24 of 34), so without is mostly short solo episodes that retain well on their own.	24	2.55% vs 2.4%	7.65% vs 11%	Neutral
Number in the titleInside the noise floor.	12	2.5% vs 2.6%	8.8% vs 7.8%	Neutral
Negation or loss wordNot a lever either way.	14	2.55% vs 2.4%	7.2% vs 8.8%	Neutral
Dollar figuren=2 is too small to act on, and the compliance gate keeps comp and valuation dollar figures out of packaging anyway.	2	2.05% vs 2.5%	9.95% vs 7.65%	Test
Authority or insider wordDirection is still negative. The old catalog had already nearly eliminated this trigger, from 6 titles down to 1.	1	1.8% vs 2.5%	6% vs 8.4%	Avoid

Through the HPM hook formulas

Every title tagged with its primary HPM hook formula (from the Hook Development Playbook). Same two signals.

Story

2.8% click-through · 17.4% viewed · n=5

Now the strongest formula on the channel: above-median click and the best retention. Upgraded from 'keeps the room at an average click.'

He Built 75% of His Book Eating Lunch, The Free Lunch Strategy | Victor Crumity Jr

Before/After Contrast

2.7% click-through · 9.95% viewed · n=8

Still ahead on both signals, and strengthened: n moved from 5 to 8.

He Drove Uber at Night to Build His Book by Day | Phil Coffin

Insider Truth

2.5% click-through · 6.6% viewed · n=8

Flat click, and retention has slipped below the channel median.

How to Spot a Bad FMO Before It Ends a Career | Ron Landes

Pattern Interrupt

2.4% click-through · 4.6% viewed · n=7

Lost its click edge (was 3.0, the Season 3 house style); now at or below median, and still the weakest retention on the board.

What the Clinic Actually Sees When an Agent Walks In | Brian Clemens

Statistical Shock

2.5% click-through · 4.8% viewed · n=3

Flat click, weakest retention; n is still small.

The Industry Takes 40 Days to Ready an Agent, She Did It in 2 | Jessica Curtis

Villain Identification

1.9% click-through · 9.9% viewed · n=3

Still costs the click.

I Was the Gas Pedal With No Brakes, The Partner Who Saved the Agency | Jonathan Manotti

Thumbnail triggers

All 38 thumbnails share one system: brush-lettered white headline, orange accent underline, dark studio, the Closed! mark top left, full contrast on every image. Those never vary, so they cannot be tested. They are the brand and they stay. Everything below is what varies inside that system.

Trigger	n	Click-through, with vs without	Percent viewed, with vs without	Verdict
Digit, $ or % in the headlineConfirmed. Numbers belong in the title, not the picture.	5	1.8% vs 2.6%	6.4% vs 9.9%	Avoid
Host pointing or gesturingConfirmed, unchanged.	10	2.9% vs 2.35%	7.45% vs 9.15%	Use
Two faces: host plus guestSlightly weaker click-through lift than the last read (0.3 points), but it still holds.	25	2.6% vs 2.3%	6.4% vs 12.6%	Use
Second person to the viewer (YOU'RE THE BAD LEAD)The picture may confront the viewer even though the title should not.	10	2.65% vs 2.35%	11% vs 8.25%	Use
Third person about the guest (SHE NEVER BUYS LEADS)Top of the table, and strengthened. "They're" is ambiguous, so n moved from 3 to 4.	4	3.45% vs 2.4%	17.25% vs 8.95%	Use
Headline of 3 words or fewerUnchanged.	10	2.7% vs 2.4%	5.1% vs 11.25%	Test
Headline of 5 words or moreToo much to read at feed size.	4	2.1% vs 2.45%	8.25% vs 9.7%	Avoid
Flat statement, nobody in it (FAKE AGENCY OWNER)A line with a person in it still beats a line with nobody in it.	14	2.2% vs 2.55%	4.8% vs 11.25%	Avoid
First person confession (I LOST MY HOUSE)Weakened to a true neutral.	5	2.4% vs 2.5%	10.4% vs 8.1%	Neutral
Laughing or smilingFine.	11	2.5% vs 2.3%	12.1% vs 8.4%	Neutral
Surprised faceConfirmed. The shocked-face convention still does not work for this audience.	4	2% vs 2.5%	7.25% vs 8.95%	Avoid
Headline repeats the title (50 percent or more of the words)Correlation, not cause: it now tracks the best-retaining videos, but that is likely because both signals track a stronger story, not because echoing the title causes retention. Do not recommend echoing on that basis.	8	2.5% vs 2.45%	19.5% vs 6.8%	Neutral

The pairing that wins

Rank the promoted episodes by click-through and read the thumbnail line next to the title. The winners split the job: the thumbnail is the emotional line aimed at the viewer or at a person, the title is the third-person story with a name and a payoff. The losers double up: a number in both, a generic statement in both, or an accusation in the picture with nothing human in the title to resolve it.

Top four by click-through

4.4% click-through 2.8% viewed

SCARED TO SCALE

He Reads Every Slack Message in His Company, The Cost of Scaling | James Jiang

James Jiang

4% click-through 17.4% viewed

YOU'RE THE BAD LEAD

He Built 75% of His Book Eating Lunch, The Free Lunch Strategy | Victor Crumity Jr

Victor Crumity Jr

4% click-through 26.4% viewed

HE DROVE UBER AT NIGHT

He Drove Uber at Night to Build His Book by Day | Phil Coffin

Phil Coffin

3.8% click-through 5.4% viewed

THEY'RE WATCHING YOU

What the Clinic Actually Sees When an Agent Walks In | Brian Clemens

Brian Clemens

Bottom four by click-through

1.1% click-through 8.4% viewed

I WAS "OUT OF CONTROL"

I Was the Gas Pedal With No Brakes, The Partner Who Saved the Agency | Jonathan Manotti

Jonathan Manotti

1.4% click-through 4.8% viewed

2 DAYS, NOT 40

The Industry Takes 40 Days to Ready an Agent, She Did It in 2 | Jessica Curtis

Jessica Curtis

1.4% click-through 27.8% viewed

FMO SECRETS EXPOSED

How to Spot a Bad FMO Before It Ends a Career | Ron Landes

Ron Landes

1.5% click-through 9.5% viewed

TOUCH 2 = BROKE

Four Years of No, Then a $20,000 Yes, The Doctor Follow-Up | Justin Jacobs

Justin Jacobs

The Closed packaging format

Write every new episode's packaging against this before it is scheduled. Loop v2: optimized for click-through and retention only, recruitment framing is out. It sits on top of the HPM Hook Development Playbook, and the Friend Test applies to every line: would a caring friend say this?

Scoreboard: track CTR and 0:30 retention together

Track click-through rate and 0:30 retention together, per episode, not just at the bucket level.
Baseline click-through is 2.5 percent; the best episodes hold 67 to 73 percent of viewers at 0:30.
A clicky thumbnail that tanks the hook is a fail, even if the click-through number alone looks good.

Title: name last, third-person story in part two

The guest's name is now standard practice; keep it after a pipe at the end.
The live lever is a third-person story or reframe in the second clause: what happened to the guest, stated plainly, not teased.
Prefer the Story or Before/After Contrast formula. Use Pattern Interrupt only when the first 30 seconds pays it off on camera.
The old avoids are already purged from the catalog: question marks, you/your, insider words, titles of 55 characters or fewer. Keep them out.

Thumbnail: tested defaults, not laws

Host mid-gesture (pointing), a person-line rather than a flat statement, third person about the guest or second person to the viewer.
No digits. No surprised face. Four words or fewer.
These are tested defaults on a small n, not rules. Validate any one of them with a single-attribute A/B on new episodes before treating it as settled.

Cold open: earn the first 30 seconds

The guest is on screen and named by the 10-second mark.
The thumbnail's line is spoken or answered on camera before 0:30.
If the title is a Pattern Interrupt, the interrupt is demonstrated in the cold open, not restated as a slogan.

Never

Comp, override or valuation dollar figures in a title or thumbnail.
A competitor FMO's name.
A stat the transcript does not contain.
An em dash.

Season 4 reframes

Update: S4E1 through S4E3 have since published, and their live titles land close to the drafted picks below, so this section is now a historical record of what was proposed, not an open ask. The reads and proposals are kept exactly as originally written, for that record. Season 4's first three episodes were packaged before the original read. The drafted picks were strong on structure; this is what the data said about each and one proposal per episode. Proposals only, as originally written. Naming and copy were the owner's call.

S4E1

Sergio Diaz

Reverse interview: contracting and carriers

Drafted

Title: The Carrier Isn't Your Enemy. Here's How the Contracting Game Actually Works | Sergio Diaz | S4 Ep. 1
Thumbnail: NOT YOUR ENEMY
Alternate: $600 TO ONBOARD YOU

What the data says

The title carries Your (second person: 2.15 vs 2.65 click-through) and an open loop (Here's how: clicks up, retention down).
The thumbnail is right on every count: three words, second person, no digits.
The alternate thumbnail has a dollar figure (the 1.9 bucket) and is a company cost figure the compliance gate keeps out of packaging anyway.

Proposed

Title: A Carrier Asked Justin Where the Sales Were. He Said Come Back in Three Years | Sergio Diaz
Thumbnail: NOT YOUR ENEMY
First 30 seconds: Sergio in Justin's chair; the come-back-in-three-years line spoken on camera before 0:30.
Formula: Story Opening leading to Pattern Interrupt. Elements: Competence Restoration, Insider Knowledge.

S4E2

Brenda Powell

The unlikely top producer

Drafted

Title: I Bet Against Her on Day One. Now She's One of Our Top Producers | Brenda Powell | S4 Ep. 2
Thumbnail: I BET AGAINST HER
Alternate: THEY CALL HER SISTER

What the data says

The title is already the channel's best pattern: Before/After Contrast, a named guest, two parts, no you, no question. Keep it.
The thumbnail repeats the title's first clause and is first person (2.15 vs 2.5). Third-person lines about the guest top the thumbnail table (3.1).
Skip 14 POLICIES, FIRST AEP: digits.

Proposed

Title: I Bet Against Her on Day One. Now She's One of Our Top Producers | Brenda Powell
Thumbnail: SHE WASN'T THE PICK
First 30 seconds: Justin's admission on camera and Brenda's reaction by 0:20.
Formula: Before/After Contrast. Elements: Identity Protection, Hope Activation.

S4E3

Umar Khan

Write the other half of the policy

Drafted

Title: Most Agents Write Half the Policy. This 22-Year-Old Writes the Other Half | Umar Khan | S4 Ep. 3
Thumbnail: HALF THE POLICY
Alternate: 99% NEVER LEAVE

What the data says

The title is a clean Pattern Interrupt with a third-person subject and a name: the highest-click formula. Its risk is retention, so the interrupt has to be paid off early.
The thumbnail is a flat statement with nobody in it (2.25 vs 2.6) and repeats the title.
99% NEVER LEAVE is a digit thumbnail (the 1.9 bucket).

Proposed

Title: Most Agents Write Half the Policy. This 22-Year-Old Writes the Other Half | Umar Khan
Thumbnail: YOU ONLY WROTE HALF
First 30 seconds: Umar on screen; the naked-plan line and the co-pay example before 0:30, so the interrupt lands as mechanism, not slogan.
Formula: Pattern Interrupt with Competence Restoration and Hope Activation.

Fix list

Housekeeping the read surfaced.

Done: question-mark and second-person titles are purged.

The 2026-09-02 mass retitle removed all three question-mark titles and every second-person (you/your) title from the catalog. Both buckets are confirmed at n=0 against the 2026-09-18 rejoin.

Retest the digit thumbnails before re-promotion.

The digit, dollar or percent headline bucket (n=5) is still avoid, confirmed unchanged (1.8 vs 2.6 percent click-through). Test a person-line variant before any of those episodes is re-promoted.

Tag the two newest episodes' thumbnails.

Brenda Powell (NYx7PXPpwOE) and Sergio Diaz (5h1_dUgXbOk) are promoted with lifetime CTR but have no thumbnail tag yet, so they are excluded from the thumbnail-trigger table above. Tag both against the existing attribute set.

Check the Sergio Diaz title's missing pipe delimiter.

Video 5h1_dUgXbOk's current title is missing its pipe delimiter before the guest name in Studio. Confirm and fix the formatting so it parses consistently with the rest of the catalog.

Declare the thumbnail, title and formula pairing in each new episode file.

Thumbnail line, title, formula, and elements, filed in the episode's own file. More data points with declared intent make the next read of this page a test instead of a survey.