Amazon's Second Phone Call: Inside 'Project Transformer' and the Billion-Dollar Bet to Re-Enter the Smartphone Market

Amazon is building a smartphone again. More than a decade after the Fire Phone became one of the most prominent hardware failures in tech history, the company has quietly assembled a team to develop a new device, codenamed "Project Transformer," that would put Alexa and artificial intelligence at the center of the mobile experience [1][2].

The project, first reported by Reuters on March 20, 2026, is being developed within a secretive unit called ZeroOne inside Amazon's Devices and Services division [1]. The effort raises a question that previous challengers — Microsoft, Essential, Facebook — answered with billions in losses: can any company break into a smartphone market dominated by Apple and Samsung?

The Fire Phone's Wreckage

To understand why this matters, start with Amazon's first attempt. The Fire Phone launched in June 2014 at $649 unlocked, exclusive to AT&T in the United States. It ran Amazon's proprietary Fire OS, which lacked access to the Google Play Store, and featured a complicated multi-camera system called "Dynamic Perspective" that rendered 3D images but drained the battery and caused overheating [3][4].

Sales were immediately poor. Estimates suggest Amazon sold fewer than 35,000 units in the first 20 days [3][5]. By September 2014, the price had been slashed to $0.99 with a two-year contract. In October, Amazon took a $170 million write-down, including $83 million for unsold inventory [6][7]. The company recorded a $544 million operating loss that quarter — the largest in its history at the time — on $20.58 billion in revenue [6].

By August 2015, the Fire Phone was dead. Amazon never disclosed total development costs, but analysts estimated the full investment at several hundred million dollars beyond the write-down, factoring in years of R&D at Amazon's Lab126 hardware division [3][4].

The failure was widely attributed to Jeff Bezos's personal insistence on the Dynamic Perspective feature, which engineers considered a gimmick, and to a product strategy that prioritized Amazon's shopping interests over what consumers actually wanted [4][5].

What's Different This Time: AI, Satellites, and a New Team

Project Transformer is being led by J Allard, a former Microsoft executive who co-created the Xbox and led the Zune music player project [1][2]. Allard joined Amazon in 2025 to head ZeroOne, a special projects team whose mandate is to create "breakthrough consumer product categories" [2].

Allard reports up through Panos Panay, the former Microsoft Surface chief who took over Amazon's entire Devices and Services division in late 2023, replacing longtime leader Dave Limp [8][9]. The appointment of two former Microsoft hardware executives to oversee a smartphone project carries a certain irony, given Microsoft's own catastrophic foray into mobile.

Several technical capabilities distinguish this effort from the Fire Phone era:

AI-native design. Rather than building around a traditional app store, Amazon reportedly envisions the phone using Alexa+ — the generative AI-powered assistant launched in February 2025 — to handle tasks that currently require separate apps [10][11]. Users could theoretically order an Uber, buy groceries, or book a repairman through Alexa without downloading individual applications. Amazon has reported "tens of millions" of Alexa+ sign-ups with engagement rates two to three times higher than the original Alexa [12].

Satellite connectivity. Amazon's Project Kuiper satellite constellation could provide direct-to-device connectivity, eliminating dead zones without relying on traditional cell towers [10]. The company has signed backhaul agreements with carriers including Verizon and Vodafone [10]. This is infrastructure the Fire Phone never had access to.

A custom lightweight OS. Reports suggest the device may run a React Native-based operating system called "Vega OS," a Linux-based system designed for tighter ecosystem control than the forked Android approach that hamstrung the Fire Phone [10].

Two-tier product strategy. The development team is considering both a traditional smartphone and a minimalist "dumbphone" inspired by the Light Phone, targeting the growing screen-addiction backlash [2][12]. Positioning Transformer as a secondary device rather than a primary iPhone or Samsung replacement would lower the competitive bar considerably.

A Graveyard of Challengers

Amazon would enter a market littered with the wreckage of well-funded competitors who tried and failed to break the Apple-Samsung duopoly.

Microsoft and Nokia. Microsoft acquired Nokia's Devices and Services division in April 2014 for approximately $7.2 billion. By July 2015, CEO Satya Nadella wrote off $7.6 billion — more than the purchase price — and laid off 7,800 employees [13][14]. Windows Phone peaked at roughly 3.6% global market share before Microsoft abandoned it entirely. Nadella subsequently redirected the company toward cloud computing and cross-platform software, releasing Office and Outlook on iOS and Android [14].

Essential Phone. Andy Rubin, the creator of Android itself, launched Essential Products in 2017 with $330 million in funding and a $1 billion valuation. The Essential Phone sold approximately 88,000 units in its first year, with only about 5,000 in initial pre-sales [15][16]. The company shut down in February 2020 without ever shipping a second product [16].

Facebook and HTC. In April 2013, Facebook partnered with HTC to release the HTC First, a phone pre-loaded with the "Facebook Home" interface. AT&T reportedly sold fewer than 15,000 units before slashing the price from $99 to $0.99 within a month [17][18]. Facebook later invested billions in its Building 8 hardware lab, which produced the Portal video-calling device before that too was discontinued [17].

Humane and Rabbit. More recently, AI-native hardware startups Humane (AI Pin) and Rabbit (R1) both attempted to bypass the smartphone paradigm entirely. Both received poor reviews and were quickly discontinued [12].

The pattern is consistent: deep pockets, strong brands, and technical talent have not been sufficient to dislodge incumbents whose advantage compounds through app ecosystems, carrier relationships, and consumer habit.

The Market Amazon Would Enter

Source: Counterpoint Research

Data as of Jan 12, 2026CSV

The timing presents additional headwinds. IDC forecasts global smartphone shipments will fall 12.9% to 1.12 billion units in 2026 — the largest annual decline on record [19][20]. A global memory chip shortage, driven by AI data center buildouts consuming DRAM and NAND supply, is pushing the average smartphone selling price up 14% to a record $523 [19][20].

In the United States, Apple and Samsung together account for more than 75% of smartphone unit sales [21]. Apple led the global market in 2025 with 20% share, followed by Samsung at 19% and Xiaomi at 13% [22]. Even capturing 3-5% of US market share — a modest target — would require selling roughly 4 to 7 million units annually, based on an estimated US market of approximately 140 million units per year [21].

That volume would represent a significant manufacturing challenge. Amazon's hardware track record with Echo and Kindle shows it can produce consumer electronics at scale — the company has sold over 500 million Alexa-enabled devices worldwide [11]. But smartphones involve far more complex supply chains: display panels, cellular modems, camera modules, and carrier certification processes that Amazon has not managed since the Fire Phone.

The Opportunity Cost Question

Amazon's high-margin businesses raise a pointed question about capital allocation. AWS generated $35.6 billion in Q4 2025 revenue with a 35% operating margin, growing 24% year over year [23]. Advertising revenue hit $21.3 billion in the same quarter, up 22% [23]. Together, these segments generate operating profits that dwarf Amazon's entire devices division.

Source: Amazon SEC filings / Insider estimates

Data as of Feb 6, 2026CSV

That devices division has been a persistent money pit. Amazon's hardware business reportedly lost $25 billion between 2017 and 2021, roughly $5 billion annually [24][25]. The Alexa division alone lost $5 billion in 2022 [24]. Two rounds of layoffs in 2022 and 2023 cut headcount significantly, and key executives departed before Panay's arrival [25].

Amazon's total planned technology investment for 2026 is approximately $200 billion, spanning AWS data centers, Kuiper satellites, and device development [10]. The smartphone project's specific budget has not been disclosed, and Reuters sources noted it could be scrapped if priorities shift [1][2].

Any dollar spent on smartphone development is a dollar not spent on AWS infrastructure, where Amazon faces fierce competition from Microsoft Azure and Google Cloud. It is also a dollar not spent on advertising technology, Amazon's fastest-growing profit engine. The strategic logic must therefore rest not on the phone generating direct hardware profits — Amazon has never prioritized that with its devices — but on driving incremental spending across the Amazon ecosystem.

The Bull Case: Ecosystem as Moat

Defenders of the strategy point to several advantages Amazon holds that previous challengers did not.

Amazon has 200 million Prime members globally [11]. If even a small fraction adopted a Transformer device as a secondary phone or as their primary device at a subsidized price point, the installed base could exceed what Essential or the HTC First ever achieved. Amazon has historically sold hardware at or below cost to drive services revenue — the Kindle and Echo models both follow this pattern.

The AI angle is also substantively different from anything available in 2014. If Alexa+ can reliably replace the need to download and manage individual apps — handling commerce, entertainment, communication, and information through a single conversational interface — the traditional app store disadvantage becomes less relevant. Amazon would not need to convince developers to build for a third platform; it would need Alexa+ to be good enough that users do not miss them.

Kuiper satellite connectivity offers a genuine differentiator. No current smartphone offers built-in satellite broadband (Apple's Emergency SOS via satellite is narrowband and limited to text). A phone that works reliably in rural areas, on flights, or during natural disasters has a real value proposition, particularly for Amazon's delivery and logistics workforce.

The Bear Case: History Rhymes

Skeptics counter that these advantages are theoretical while the barriers are proven. The app ecosystem problem has only intensified since 2014. Consumers now have years of purchased apps, stored passwords, health data, and payment credentials locked into iOS or Android. Switching costs are higher than ever.

The "secondary device" positioning, while lowering expectations, also limits the addressable market. Few consumers want to carry two phones. The dumbphone trend, while culturally visible, represents a tiny market segment — Light Phone has sold in the low hundreds of thousands of units total.

Manufacturing partnerships remain an open question. Amazon would need to secure display panels, processors, and cellular modems from suppliers who prioritize Apple and Samsung orders. The 2026 memory shortage makes component sourcing even more competitive [19].

And the historical record is unambiguous. Every major tech company that has attempted to enter the smartphone market since 2013 has either failed outright or retreated. Microsoft wrote off more than it paid. Andy Rubin, who literally invented Android, could not make it work. Facebook abandoned hardware entirely. The market has consolidated, not opened.

What Comes Next

Amazon plans to launch a limited beta of the Transformer device to its "Amazon Vine" product reviewers and select Prime members by late 2026 [1][10]. The company has not confirmed the project's existence publicly.

The smartphone market in 2026 is structurally different from 2014 — both in ways that help Amazon (AI capabilities, satellite infrastructure, a mature Prime ecosystem) and in ways that hurt (deeper entrenchment by Apple and Samsung, a shrinking global market, a memory chip crisis driving up costs).

Whether Project Transformer succeeds may ultimately depend on a more fundamental question: is Amazon building a phone because it has identified a genuine gap in the market, or because the company's institutional ambition demands another attempt at the one consumer device category that has eluded it? The Fire Phone answered that question one way. Amazon is betting that 12 years of technological change — and organizational learning — will produce a different answer.