retti.aiTeardowns › I Built MrBeast and iShowSpeed Dream Gaming Rooms!
Predicted Retention Teardown

I Built MrBeast and iShowSpeed Dream Gaming Rooms!

By Matthew Beem · Gaming · 6.4M views · 36:34

I Built MrBeast and iShowSpeed Dream Gaming Rooms!

The teardown in brief

What's working

What's costing attention

The first 30 seconds

Today, I'm recreating five streamers' gaming rooms. This is insane. And I'm doing it to help Mr. Beast. >> I want to prank the internet and have all the streamers go live in identical replicas of their rooms in the studio. I need you to build them. But don't worry, I know what you're thinking. Here you go. Uh I had my

Hook fires at 0:05 with MrBeast's voice and a bag of cash — the premise is crystal clear within 22 seconds. Strong tier-1 delivery that minimizes the packaging drop to the low end of the gaming format range.

Where viewers drop

15:00 — Repetitive Room-Build Loop (Minutes 15-24) (critical)

You cycle through Pokimane's room and then Ludwig's room using the exact same structural pattern: walls go up, carpet drops, items get placed, creator sits in chair for selfie comparison shot. After doing this same loop for Speed, xQc, and Quackity earlier in the video, by the time you hit room four and five in this stretch the viewer can predict every beat before it happens. There's no new obstacle, no time pressure reintroduced, and no consequence reminder for nine straight minutes.

Why it matters — A viewer who's been watching for 15 minutes already knows exactly what each room segment looks like. When room four starts, their brain goes 'I've seen this three times' — and they don't need to see it a fourth.

0:00 — Missing Explicit Stakes — What Happens If The Prank Fails? (critical)

The video never tells the viewer what's actually at risk. 'Fool chat' is mentioned a few times but there's no consequence if they don't fool anyone. Will MrBeast cancel the prank? Will the streamers be embarrassed? Will money be lost? Without a stated consequence, every close-up comparison shot ('this has to be perfect') feels like craft pride rather than genuine tension.

Why it matters — When a viewer doesn't know what failure looks like, they don't feel fear — and fear is what keeps people from clicking away during a 36-minute build video.

6:46 — Sponsor Interrupts At Peak Momentum (6:46 and 11:11) (moderate)

The first sponsor read hits at 6:46 right after a satisfying Luffy sculpt reveal moment, stalling the xQc-arriving-soon urgency you just planted. The second sponsor read and subscribe CTA at 11:11 interrupts right before the Quackity 10-minute challenge, which is one of the video's best tension mechanics. Both reads feel like a hard stop rather than a bridge.

Why it matters — You've built time pressure twice ('xQc arriving soon,' 'we have to do Quackity in 10 minutes') and both times a sponsor read lets the viewer breathe right when they should be holding their breath.

0:00 — Audio Monotony — 80% Shouting For 36 Minutes (moderate)

The audio energy data shows 80% of this video at VERY_LOUD (shouting/intense) with zero quiet or normal moments beyond a single 3-second window at 5:27. In a gaming/entertainment niche this energy level is expected, but with no contrast, the moments that should feel BIG — like Lacy walking into his room for the first time at 29:19 or Speed's reaction at 31:03 — don't actually hit harder than anything else. Everything is at max volume, so nothing is max volume.

Why it matters — When the Lacy reveal and the Speed reveal and the toilet paper bed test and the wall smashing all sound identical, the climax of the video doesn't feel like a climax — it just feels like more of the same.

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