retti.aiTeardowns › I Turned My Bedroom Into A Fish Tank!
Predicted Retention Teardown

I Turned My Bedroom Into A Fish Tank!

By Matthew Beem · Gaming · 9.5M views · 16:00

I Turned My Bedroom Into A Fish Tank!

The teardown in brief

What's working

What's costing attention

The first 30 seconds

Today, I'm turning my little sister's room into a fish tank. I'm also blowing up an entire village, completely filling a school classroom with snow. But before we see what happens when you fill a bedroom with water, we'll be testing three other insane experiments. Starting with crushing the store with giant meteors. We

Hook fires immediately — bedroom fish tank is teased in the first sentence, all four experiments are laid out in 18 seconds, and concept is crystal clear by 0:08. Strong Tier 1 delivery that minimizes the mandatory packaging drop.

Where viewers drop

0:00 — Zero Stakes — No Consequences Ever Stated (critical)

You never tell the viewer what happens if you stop or fail. The whole video is just 'watch cool things happen' with no tension — there's no reason to stay for the fish tank specifically versus clicking away after the meteor.

Why it matters — Without a consequence, every experiment feels like a standalone clip. Viewers can leave after the TNT explosion feeling satisfied — you gave them no reason to stay for the main event.

0:00 — Sustained Maximum Audio Energy — Zero Contrast (critical)

You are shouting at maximum intensity for 98% of this video. The audio data shows 87% of the runtime at VERY_LOUD (-6 to -10dB) with zero normal or quiet sections. When everything is a 10, nothing is a 10 — your biggest moments (meteor impact, TNT explosion, sister arriving) land the same as talking about where to place a toy figure.

Why it matters — Sustained maximum energy for 16 minutes is audiologically exhausting for young viewers. The loudest, most exciting moments get no emotional contrast to make them pop. Your sister walking in at 11:13 — the most genuine moment in the video — sounds exactly like every other second.

1:53 — Formulaic Experiment Structure — Predictable After Experiment 1 (moderate)

After the meteor segment, viewers can predict exactly how each experiment plays out: (1) explain the Brawl Stars tie-in, (2) start small, (3) escalate, (4) big destruction/chaos. The TNT village and snow classroom follow this arc so precisely that by experiment 3, there's nothing unexpected.

Why it matters — Predictable structure removes the single most powerful retention mechanic you have: surprise. Viewers who've pattern-matched the format stop actively watching and start passively waiting for the next destruction clip.

8:52 — Sponsor Read Mid-Momentum (moderate)

At 8:51, after the snow experiment wraps, you deliver a 33-second Brawl Stars download CTA. This placement is between experiments rather than mid-tension, which is its best feature — but the hard cut back to 'it's finally time to see what happens when you flood your sister's room' at 9:06 loses the momentum the snow segment built.

Why it matters — You gave viewers a natural exit point right before your most-anticipated segment (the fish tank bedroom). The viewer who was only there for the snow experiment just got told the video is pausing — and they left.

How the video is built

What any creator can steal

More teardowns from Matthew Beem

Want this on your own video?

Paste any YouTube URL and Retti maps every drop, spike and plateau to the moment that caused it.

Analyse a video free