Case Study: The Loop Architect

"I used to think hooks were the hard part. Turns out, hooks are easy. Keeping people watching is the real architecture."

Overview

This case study follows Tomás Herrera, a 17-year-old history creator on TikTok, as he transforms his content from single-hook videos with sharp dropoff curves into multi-loop structures that sustain attention from first second to last. His journey illustrates the practical difference between opening a curiosity gap and maintaining one — and introduces a reusable framework for "loop architecture."

Skills Applied: - Open and closed loops (Zeigarnik effect) - Loop timing and nesting - Micro-satisfaction spacing - Mystery, suspense, and dramatic irony as structural tools - The "always one loop open" principle

Part 1: The Hook-and-Cliff Problem

Tomás's Starting Point

Tomás made 60-second history videos — "wild true stories from history that nobody talks about." His hooks were excellent. Opening lines like "In 1932, Australia lost a war to birds. Actual birds." and "A prisoner escaped from Alcatraz using a vacuum cleaner motor and a raincoat" consistently stopped the scroll.

His analytics told a clear story:

TYPICAL RETENTION CURVE (before)

100% |████
 80% |████████
 60% |██████████
 40% |████████████████
 20% |████████████████████████████████████████████████
  0% |————————————————————————————————————————————————————
     0s    10s    20s    30s    40s    50s    60s

Strong at 0-10 seconds (the hook worked), then a steep, steady decline. By 30 seconds, he'd lost 60% of viewers. By the end, only 20% remained.

"I kept trying to write better hooks," Tomás said. "But the hooks were already good — that wasn't the problem. The problem was everything that came AFTER the hook."

The Diagnosis

When Tomás mapped his loop structure, the issue was obvious:

Typical video — loop architecture:

0s ——— 10s ——— 20s ——— 30s ——— 40s ——— 50s ——— 60s
[MAIN LOOP: "Australia lost a war to birds" ———————————— CLOSES at 58s]
                    (no other loops open)

One loop. Opened at the start, closed at the end. Everything in between was straight narration — Tomás telling the story linearly, fact by fact, in chronological order.

The problem: the main loop ("What happened in the Emu War?") had a moderate curiosity gap — interesting enough to hook, but not agonizing enough to sustain 60 seconds of attention on its own. By second 20, many viewers had gotten "enough" of the answer from context clues to satisfy their curiosity partially. They'd grasped the gist (Australia sent soldiers after emus and it didn't go well), and the remaining details didn't seem worth another 40 seconds.

Tomás was asking one question and taking 60 seconds to answer it. In a feed environment, that's an eternity.

Part 2: The Loop Architecture Redesign

The Principle: One Loop Is Never Enough

Tomás's mentor (a history YouTuber with 2M subscribers) told him: "A great video isn't one big question. It's a relay race of questions. Each one hands off to the next before the runner stops."

This reframed Tomás's entire approach. Instead of opening one gap and filling it linearly, he needed to architect sequences of gaps that kept the Zeigarnik tension refreshed throughout the video.

The Framework: Loop Architecture Template

Tomás developed a template he called the "5-Loop 60":

LOOP 1: THE HOOK (0-3s)
Opens: The main question/mystery
Type: Usually informational surprise or mystery
Example: "Australia declared war on birds."

LOOP 2: THE COMPLICATION (5-12s)
Opens: A secondary question that deepens or complicates the main one
Type: "But here's the weird part..."
Example: "They sent soldiers with machine guns. The emus won."
[Closes partially: LOOP 1 gets a micro-answer — yes, it really happened]

LOOP 3: THE HUMAN ELEMENT (12-25s)
Opens: A personal/emotional question about someone involved
Type: "And one person's story makes it even stranger..."
Example: "The commanding officer's report to Parliament is the funniest
military document ever written."
[Closes: LOOP 2 — explains HOW the emus won (speed, dispersal, thick feathers)]

LOOP 4: THE TWIST (25-40s)
Opens: Something that reframes the entire story
Type: "But here's what nobody mentions about the Emu War..."
Example: "The farmers who started the whole thing? They asked for
ANOTHER war. Twice."
[Closes: LOOP 3 — reads the actual report excerpt]

LOOP 5: THE KICKER (40-55s)
Opens: A final question or ironic connection
Type: "And the wildest part is what happened next..."
Example: "Australia's solution? They gave up on war and put up fences.
The fences worked. Sometimes the boring answer is the right one."
[Closes: LOOP 4 — explains the second and third attempts]
[Closes: LOOP 1 — complete story now told]
[Closes: LOOP 5 — the ironic kicker resolves]

TAG (55-60s): Optional serial hook or emotional landing

The Redesigned Emu War Video

Here's how Tomás restructured his most popular topic:

Results

The retention curve transformed:

TYPICAL RETENTION CURVE (after)

100% |████
 90% |██████████
 80% |████████████████
 70% |████████████████████████
 60% |██████████████████████████████████
 50% |██████████████████████████████████████████████
  0% |————————————————————————————————————————————————————
     0s    10s    20s    30s    40s    50s    60s

The steep cliff became a gradual slope. More importantly, there were small increases in retention at the 12s, 28s, and 45s marks — exactly where new loops opened. Each new question created a small surge of re-engagement.

Part 3: The Architecture Applied to Other Niches

Tomás shared his 5-Loop framework with three friends, each in a different niche. Here's how they adapted it:

Recipe Video (Cooking Niche)

Room Tour (Lifestyle Niche)

Fitness Progress (Fitness Niche)

Part 4: Common Loop Architecture Mistakes

Based on Tomás's experimentation and feedback from his friends, five common mistakes emerge:

Mistake 1: All Loops Open at the Same Time

Opening 3-4 questions in the first 5 seconds creates confusion, not curiosity. The viewer can't track multiple gaps simultaneously. Solution: stagger loop openings.

Mistake 2: Loops Close in Order

If Loop 1 closes, then Loop 2, then Loop 3, the video feels predictable. Solution: close loops out of order — resolve the second question before the first, creating unexpected satisfaction patterns.

Mistake 3: No Micro-Satisfactions Between Loops

A new loop opening is not a satisfaction — it's a new question. If loops open without any closing in between, the viewer accumulates frustration rather than engagement. Solution: always close (or partially close) a loop before or simultaneously with opening a new one.

Mistake 4: The Main Loop Is Too Weak

If the main loop (the overarching question) isn't compelling enough, no number of mini-loops will save the video. The architecture amplifies the main question — it can't replace it. Solution: invest the most creative energy in the main gap.

Mistake 5: The Serial Hook Is Stronger Than the Video

If the tease for the NEXT video is more interesting than the video the viewer just watched, it creates resentment rather than anticipation. Solution: the serial hook should feel like a natural next step, not a better video you're withholding.

Discussion Questions

1. Tomás's retention curve showed small INCREASES at the points where new loops opened. How does this relate to the orienting response (Chapter 1) and prediction error (Chapter 4)? Is each new loop functioning as a "pattern interrupt" for attention?

2. The 5-Loop framework is designed for 60-second videos. How would you adapt it for (a) a 15-second video and (b) a 10-minute video? What's the minimum number of loops that sustains attention?

3. Tomás's serial hook ("Liechtenstein sent 80 soldiers and came back with 81") is arguably more intriguing than the Emu War story. Is this a violation of Mistake 5? Or does it work because the current video was satisfying enough?

4. Compare Tomás's approach (many small loops) with Luna's "Reverse Process" series from Section 5.3 (one sustained mystery). Are there content types where a single, strong loop is better than multiple loops? What determines which approach works better?

Your Turn: Mini-Project

Option A: Take one of your existing video ideas and apply the 5-Loop 60 template. Map each loop: what opens, what type of curiosity it uses, when it closes, and what micro-satisfaction it provides.

Option B: Record your retention analytics on your three most recent videos. Identify the exact timestamps where viewers leave. For each dropoff point, diagnose: was there a loop closing without a new one opening? Was there a satisfaction desert? Propose specific loop architecture fixes.

Option C: Apply the 5-Loop framework to a non-video context: a 5-minute school presentation, a college essay, or a conversation where you're trying to convince someone of something. How does loop architecture translate to live communication?

References

• Note: Tomás Herrera is a composite character based on real creator experiences. The Emu War is a real historical event (the "Great Emu War" of 1932 in Western Australia). The military quote is attributed to Major G.P.W. Meredith. Metrics are illustrative of documented patterns in loop-based content restructuring.