Report: Amazon's hardware group is on track to lose $10B in 2022, as many Echo devices are sold at cost and plans to monetize Alexa have failed
Layoffs reportedly hit the Alexa team hard as the company's biggest money loser. — Amazon is going through the biggest layoffs …
Ars TechnicaRon Amadeo
Context & Ripple Effects
This report lands mid-purge: Amazon had just opened a cost-cutting review of its unprofitable businesses after the devices unit posted a $5B+ operating loss in recent years, and CEO Andy Jassy had already warned the layoffs would run into 2023 following cuts across Devices and Books. The $10B figure reframes those moves from trimming to triage — Alexa is not one underperforming line among several but the company's biggest money loser.
The deeper arc is older than this news cycle: Amazon scaled back its Lab126 hardware center after the Fire Phone failure in 2015, then kept selling Echo at cost on the bet that voice commerce would pay for the hardware. A later accounting put the cumulative damage at $25B+ lost on devices between 2017 and 2021, confirming the monetize-it-later model never closed the gap.
First-order effects
The Alexa team absorbs the deepest cuts in Amazon's planned ~10K-role reduction, with layoffs beginning across the Alexa, Kindle, and Halo hardware teams within weeks of this report.
Second-order effects
Echo's at-cost pricing becomes untenable once the services revenue meant to subsidize it fails to materialize, forcing either higher device prices or a thinner, cheaper Alexa feature set — and giving rivals like Google an opening to define what consumers will pay for voice assistants.
Third-order effects
If the pattern holds, the 'sell hardware at a loss, monetize the service' playbook loses credibility across consumer tech, pushing vendors toward subscription-priced assistants or premium hardware margins instead of subsidized volume.
The trend: Voice-assistant economics are collapsing under their own subsidy model, forcing Big Tech to treat AI hardware as a business that must stand on its own rather than a loss leader for services.
The BI report spoke w/ “a dozen current and former employees on the company's hardware team,” who described a division in crisis. Just about every plan to monetize Alexa has failed, w/ one former employee calling Alexa “a colossal failure of imagination.” https://arstechnica.com/…
I've been predicting this since day 4. Most of this time everyone has laughed at me. Smart speakers were never really bought. They were sold. https://arstechnica.com/...
by year four of the Alexa experiment, “Alexa was getting a billion interactions a week, but most of those conversations were trivial commands to play music or ask about the weather.” Those questions aren't monetizable. Apple's Siri was at 4 billion interactions a week in Aug 2019
these sold over 70 million units. privacy advocacy over what should be a no-brainer (the world's largest company listening to everything happening in your house) was useless compared to brute economic facts (they couldn't monetize user commands) https://twitter.com/...
Voice was a gamble. It's a low-bandwidth interface, so it needs to understand a huge amount of context to be able to do what you are intending to ask for. And as we've seen, AI has not lived up to the hopes for that. https://twitter.com/...
Giant microphone and so-so speaker as centerpiece of home life. No thanks. Amazon Alexa is a “colossal failure,” on pace to lose $10 billion this year | Ars Technica https://arstechnica.com/...
The nominal convenience in search gained by speech was never enough to justify getting an Echo — and most of the “skills” are nearly worthless. As many have said to me, the devices just become expensive kitchen timers https://arstechnica.com/...
The gizmo is very useful but cannot be monetized. People use it to turn the lights on and off and ask how much a quarter pound of butter weighs in grams, but no one wants to use it to buy stuff. Sounds familiar. https://twitter.com/...
Did AI fail to meet expectations again? Personally, I would pay real money for a virtual assistant of the caliber of the OS in the film Her, but not for an AI with the narrowness of Alexa or the hallucinations of Galactica. https://twitter.com/...
having worked in the AR space on a product that was voice command only, what we learned was that controlling computers by voice alone is inherently kind of weird and no one wants to do it for anything other than trivial commands. we lost a lot less money than this, though. https:…