CASE STUDY · QUIBI

We ran a $1.75 billion failure through Assembly.

In 2020, Quibi raised $1.75 billion, assembled a Hollywood and Silicon Valley dream team, and launched the most anticipated app in streaming. Six months later it shut down. We rebuilt its pre-launch pitch as a product brief, ran it through a synthetic society, and the personas rejected it for the same reasons the real market did.

93%
of the society landed resistant
5 of 5
documented failure factors surfaced
6 mo
from launch to Quibi's shutdown
01 / The company

A $1.75 billion bet that lasted six months.

Quibi launched in April 2020 with $1.75 billion in funding, a sum almost no consumer product raises before shipping. It was founded by Jeffrey Katzenberg, the DreamWorks co-founder, and run by Meg Whitman, the former CEO of eBay and HP. The product was premium, studio-quality shows cut into "quick bites" under ten minutes, built only for the phone, with a signature feature called Turnstyle that reframed each scene as you rotated the device. Pricing started at $4.99 a month behind a 90-day free trial.

The launch drew roughly 1.5 million downloads, but paying subscribers stalled near 500,000 against a first-year target of 7 million. Most users did not stay once the free trial ended. In October 2020, six months after launch, Quibi announced it was shutting down. Roku later acquired its entire content library for less than $100 million.

Every post-mortem since has converged on the same short list of reasons it failed. We wanted to see whether a synthetic society, handed only the pre-launch brief, would arrive at that same list on its own.

02 / The test

We gave Assembly the brief, not the verdict.

We reconstructed Quibi's offering as a standard Assembly product brief, the same input any founder submits: format, platform, price, features, and audience. We renamed it Snackable, removed everything that identified the company, the team, or the year, and ran the simulation.

How we ran it

  • 01Reconstructed Quibi's offering as a standard product brief.
  • 02Renamed the product Snackable and described it only by its attributes.
  • 03Gave Assembly only what a founder knows on day one: price, format, platform, features, audience.
  • 04Left out the outcome, the funding, the team, and every post-mortem.
Assembly never knew it was Quibi.
03 / What we fed Assembly

The product brief

Every field below is an attribute a founder would know on day one. Nothing in it reveals the company, the year, or what happened next.

Product
Snackable
One-liner
Premium, studio-quality shows in episodes under ten minutes, made for your phone.
Format
Professionally produced scripted series, reality, and news, each cut into 5 to 10 minute "quick bites." No user-generated content.
Platform
Mobile only at launch. No TV, no desktop. Signature rotate-to-reframe playback that reshoots each orientation as its own intentional shot.
Constraints
Content cannot be clipped, screenshotted, or shared to other platforms.
Pricing
$4.99/mo with ads, $7.99/mo ad-free. 90-day free trial.
Target
Young professionals 25 to 35. Commuters and on-the-go millennials who watch video in short gaps during the day.
Competes with
YouTubeTikTokNetflixInstagramwatching nothing
04 / What Assembly found

A society of 30 debated it. Then it landed.

After four rounds of debate, the personas settled into a private stance. The result was not close.

Where the society landedFinal opinion · after the debate
0
Receptive
2
Uncertain
28
Resistant
The synthetic society
Each node is a synthetic persona. Edges connect personas who debated in the same room. Color is final stance.
Receptive Uncertain Resistant
What they pushed back on
Synthetic objections, ordered by how often they came up across the debate.
Price-to-value concern93%
Competitor already solves it7%
What would change their minds
Synthetic proof needs, ordered by how much they would shift the room.
A head-to-head comparison13%
What they asked for
"Evidence the content is compelling enough that people actually keep watching after the free trial."
"Show me celebrity-driven episodes good enough that I'd choose them over scrolling, and prove the rotate format actually adds something beyond novelty."
Simulated intent
The next action each persona expressed inside the simulation.
Would compare to current option14
Would reject outright13
Would consider if proven2
Loyal to a free alternative1
Hardest to convince
The segments that anchored on price-to-value and would not move.
Trust seeker4 / 4 resistant
Safety-visibility buyer3 / 3 resistant
Performance buyer2 / 4 resistant
Price skeptic2 / 4 resistant
Even the best-fit segments, convenience and format buyers, returned just one receptive persona each.
05 / In their own words

The room, talking.

These are verbatim from the run. Not ratings, not tags. The actual reasoning each synthetic persona gave for landing where they did.

Wren B. Resistant Price skeptic

"So I'm paying $4.99 a month for short episodes I can't even share or clip, when TikTok and YouTube give me endless short video for free? The rotate-the-phone trick is cute, but I'm not sure a gimmick justifies another subscription on top of what I already pay for Netflix. Honestly the thing that would kill it for me is the no-sharing part, half the fun of short video is sending it to a friend, and you've designed that out."

Oakley K. Resistant Objection focused buyer

"The rotate-to-reframe thing is a cute trick, but I'm not paying five bucks for a gimmick when TikTok and YouTube hand me endless short video for free. What really kills it for me is I can't clip or share anything, so it just sits there on my phone with no way to send a good moment to a friend. And since it's phone-only at launch, I can't even throw it on a TV when I'm actually relaxing at home, so when exactly am I supposed to use this?"

Reese R. Uncertain Format focused buyer

"The rotating reframe thing actually intrigues me as a format problem, shooting the same scene as two intentional shots is a cool design choice, and I appreciate that it's trying to do something deliberate instead of just cropping. But I'm hung up on paying for it when YouTube and TikTok give me endless short video for free, and I can't even share a clip I love with friends. If the content is genuinely premium and the orientation trick feels seamless rather than gimmicky, I'd give the trial a shot, but the no-share, phone-only setup makes me wonder how it keeps me past month one."

06 / What Assembly found vs what happened

Every objection it raised is a reason Quibi died.

Assembly surfaced these from the brief alone. The right column is what the post-mortems, and the market, recorded afterward.

What Assembly surfaced
What actually killed Quibi
Price vs freeWhy pay $4.99 when TikTok and YouTube are free and already fill my downtime?
RealityAfter the 90-day trial, users left in droves. Free platforms won the same spare minutes.
No shareabilityPaying for clips I can't send to a friend kills the one thing that makes short video spread.
RealityNo sharing meant no word-of-mouth. The growth engine for short video never started.
Gimmick, not a reasonThe rotate feature is a neat trick, but prove it adds something beyond novelty.
RealityTurnstyle was widely dismissed as a gimmick reviewers stopped noticing within minutes.
Retention past the trialShow me the content is good enough that I'd keep opening it after the free trial ends.
RealityRoughly 500K subscribers against a 7M first-year target. Almost no one converted.
Phone-only is fragileIt's mobile only, so I can't watch on a TV when I'm actually relaxing. When do I use this?
RealityA mobile-only product launched into a world locked at home with their televisions on.
Five objections, surfaced from attributes alone. Five of the factors every post-mortem names.
07 / Why this matters for you

Quibi had $1.75 billion and still couldn't see this. You don't need that much to find out.

You have an idea and a decision to make. Assembly runs your product through the same kind of society in minutes, before you build, so the objection that would have killed your launch shows up while you can still do something about it.

$1.75B
raised before launch
~500K
subscribers against a 7M target
6 mo
from launch to shutdown
93%
of Assembly's society resistant
Run yours

Don't Build In Blind.

Run your product through a synthetic society before you bet on it. Free for student founders.

Run a free simulation
assemblysimulator.com
Rehearse Reality.