Famous product failures,
and what each one teaches

Table: Famous product failures, their causes and lessons (company statements and contemporary reporting)

From New Coke's launch to the original's return

79 days

Raised by Quibi before it shut down in 2020

$1.75 billion

Google+ sessions under five seconds, per Google

90%

Samsung's estimated cost of the Note 7 problems

$5.3 billion+

Ford Edsel (1957)

Ford launched the Edsel with an enormous publicity campaign in September 1957, after years of planning for a mid-priced market. By the time it arrived, buyers were moving towards smaller, cheaper cars after a recession, and the car's styling, especially the vertical grille, divided opinion. Ford discontinued it in November 1959. The lesson is timing and drift: years of planning can leave a product designed for a market that has changed.

Sony Betamax (1975)

Sony's Betamax was widely regarded as having good picture quality, but its early cassettes recorded only about an hour, while JVC's rival VHS format, from 1976, offered longer recording, enough for a film or a football match. VHS also licensed its format widely, so more machines and more rental tapes appeared. Betamax lost the consumer market, though Sony kept making tapes until 2016. The lesson: match the job people actually want done, here, recording a whole programme.

New Coke (1985)

On 23 April 1985, Coca-Cola replaced its 99-year-old formula with a sweeter one that had beaten the original and Pepsi in taste tests of nearly 200,000 consumers¹. What the tests didn't measure was what Coke meant to people. Within weeks, the company was flooded with protest calls and letters, and on 11 July 1985, 79 days later, it announced the return of the original as Coca-Cola Classic². In behavioural terms it is loss aversion at national scale: taking something away hurts more than an equal improvement pleases.

Apple Newton (1993)

Apple's Newton MessagePad, launched in 1993, was an ambitious handheld computer whose headline feature was recognising handwriting. Early versions often got words wrong, and the mistakes were mocked in comic strips and on television. Later models improved, but the reputation stuck, and Steve Jobs ended the line in 1998. The lesson: when a product's promise rests on one feature, that feature must work reliably from day one; partial success is judged as failure.

Microsoft Bob (1995)

Microsoft Bob replaced the Windows desktop with a cartoon house of rooms, guided by animated characters, for people intimidated by computers. Experienced users found it patronising and slow, and newcomers soon outgrew it; Microsoft withdrew it within about a year. Two things survived it: the Comic Sans font, designed for its speech bubbles, and the idea of an animated assistant, which returned as Office's "Clippy". The lesson: a metaphor must reduce effort, not add a layer of decoration between people and their task (see our post on famous typefaces).

Segway (2001)

Unveiled in December 2001 after intense secrecy and hype, the self-balancing Segway was a genuine engineering achievement. But it fitted neither pavements nor roads, many cities restricted it, and at several thousand dollars it cost far more than a bicycle. It found niches in tours, security and warehouses, and production of the original model ended in July 2020. The lesson: design for the real environment, including rules and infrastructure, not an imagined one.

Google Glass (2013)

Google Glass put a small display and camera into a spectacle frame. The Explorer edition, sold from 2013 for $1,500, worked, but its everyday uses were unclear and its camera made people around the wearer uncomfortable; some bars and cinemas banned it. Google ended the consumer Explorer programme in January 2015 and shifted to business uses, ending the enterprise edition in 2023. The lesson: wearable and social products must be designed for bystanders as well as users.

Google+ (2011)

Google+ launched in 2011 with good design ideas, such as "Circles" for sharing with different groups, but people's friends were already on Facebook, and Google pushed sign-ups by tying the service to YouTube and other products. In 2018 Google said that 90% of Google+ user sessions lasted less than five seconds and announced it would shut down the consumer version, which closed in April 2019⁸. The lesson: sign-ups are not usage; measure what people do after they arrive.

Amazon Fire Phone (2014)

Amazon's Fire Phone, launched in July 2014, had "Dynamic Perspective", a 3D effect using four front cameras, and "Firefly" for identifying products to buy. Neither was a reason to leave an iPhone or Android phone, and the price matched premium rivals. Within months Amazon cut the price to 99 cents on contract, and in October 2014 it took a $170 million charge mostly related to unsold phones⁵. The lesson: novel features are not reasons; people switch for better everyday value.

Tata Nano (2009)

The Tata Nano was a real engineering feat: a four-seat car priced around ₹1 lakh at launch, meant for families riding together on a single two-wheeler. It was marketed and remembered as the "world's cheapest car", and its aspirational buyers did not want to be seen in one; early fire incidents and a delayed factory move also hurt. Ratan Tata later said that branding it the cheapest car was a mistake⁷. The lesson: positioning is part of the product; the same car described differently would have appealed to different people.

Myntra's app-only switch (2015)

In 2015 the fashion retailer Myntra shut its mobile and desktop websites to become app-only, betting that shopping would move entirely to apps. Many customers still searched, browsed and shared on the web; Myntra brought back a mobile site in early 2016 and its desktop site by mid-2016⁹. The lesson: don't remove a channel people rely on to force a habit; move them by making the new channel better.

Samsung Galaxy Note 7 (2016)

The Note 7 went on sale in August 2016 and was soon linked to battery fires. Samsung recalled it, but some replacement phones also overheated, and the company discontinued it in October 2016. In January 2017 Samsung's investigation, involving 700 researchers testing more than 200,000 devices and 30,000 batteries, blamed design and manufacturing problems in batteries from two suppliers, and the company estimated the problems would cost at least $5.3 billion⁶. The lesson: safety is the first feature, and rushed validation can wipe out every other one.

Juicero (2016)

Juicero sold a Wi-Fi-connected juice press, originally priced around $700 and later $400, that squeezed proprietary packs of chopped fruit. It raised about $120 million from investors. In April 2017 Bloomberg reporters showed the packs could be squeezed by hand almost as effectively as by the machine⁴, and the company shut down later that year. The lesson: a product must solve a real problem; elegant engineering on a non-problem is still a non-product.

Quibi (2020)

Quibi launched on 6 April 2020 with $1.75 billion raised and short, high-budget shows designed for phones. It launched as lockdowns kept people at home rather than commuting, competed with free short video on YouTube and TikTok, and at first didn't let people share clips or watch on a television. It announced in October 2020 that it would wind down³, and closed on 1 December 2020. The lesson: money can't substitute for evidence that people want the product in the form offered.

The cases above fall into a few recurring patterns, and each pattern has a cheaper test that would have revealed it.

There is a psychological trap behind many of these: the teams loved their products. Behavioural researchers call it the IKEA effect, the tendency to value what we build ourselves more than others do. It is why outside evidence, not internal enthusiasm, should decide launches.

1. Watch current behaviour — interview potential users about how they solve the problem today, and what it costs them.
2. Test demand cheaply — a landing page, a waitlist, a pre-order or a manual "concierge" version before building the full product.
3. Measure use, not opinions — track whether people come back in week two, not whether they say they like it.
4. Test in context — put prototypes into the real setting: the street, the shop, the family, the office.
5. Check positioning with buyers — test how the product is described as carefully as how it works.
6. Launch small — one city, one segment, one feature, so failure is cheap and learning is fast.

Our guide to building an MVP covers these tests in detail, and the Kano model helps sort features people expect from those that would actually delight them. If you have a product in market and want to know why people aren't sticking, a retention and churn audit finds where they drop off.

Each case is based on company statements (Coca-Cola, Google, Samsung, Amazon's earnings disclosures) and contemporary reporting, cited above. Dates in the lifespan chart are approximate launch and discontinuation dates from those records. The lessons are our analysis as product and UX practitioners. Photographs are from Wikimedia Commons and the Smithsonian's National Museum of American History, credited under each. Last reviewed October 2026.