- Amazon’s Alexa was Jeff Bezos’ pet project; now it is a target for the company’s cost reduction.
- Insider reported that the Amazon division that built Alexa is on track to lose $10 billion this year.
- Voice assistants should be revolutionary, but no one has figured out how to make them profitable.
The division that houses Amazon’s Alexa is on track to lose $10 billion this year alone, Insider’s Eugene Kim reported. And as Amazon has been hit with layoffs, the Alexa team, once 10,000 strong and a passionate project of then-CEO Jeff Bezos, is now an attractive target for cuts.
The Amazon Echo debuted in 2014 and was the company’s first real success as a hardware manufacturer. Early versions of voice assistants were already around in the wee hours, and Siri made its debut on iPhone in 2011, but Amazon’s Echo spawned a whole new class of gadget: the smart speaker. The standalone voice assistant was immediately useful, and by 2018 the company had sold more than 100 million Alexa-enabled devices.
Amazon sold the Echo device at cost to entice people to buy things on the site, but the smart speaker never became the significant revenue driver the company hoped it would. On the other hand, listening to Alexa read out two minutes of text about dishwasher pellets to make sure you’re ordering the right brand isn’t exactly a stellar user experience.
But when initial Echo sales exceeded expectations within the company, Amazon and others began considering Alexa and voice computing as a new platform. It created an app store, Alexa Skills, in hopes of sparking the same wild innovation Apple saw when developers were allowed into its app store in 2008. And it launched first-party products like microwave ovens and TVs with Alexa built-in Alexa voice assistants to other manufacturers, hoping to make the assistant ubiquitous in people’s homes.
One of the critical weaknesses for Amazon’s ambition to be in the business of almost anything is that it doesn’t have its own platform. It doesn’t control a PC operating system like Microsoft or Apple, and it doesn’t have a mobile platform like Google’s Android or Apple’s iOS (even Amazon Fire tablets only run a modified version of Android).
Voice computing, many thought, could be the next platform, and Amazon could own it.
“Glorified Clock Radio”
But Amazon’s internal data and user surveys showed that voice commands are great for a limited number of tasks: setting a timer, playing music, and getting the weather. You may have heard Alexa plaintively reminding you that it can do a lot more than just set a timer — that’s because Amazon knows most people don’t do much beyond that. Some people even stop using the product after a few weeks.
The pessimistic view was that “Amazon managed to sell a huge number of glorified clock radios,” as Benedict Evans, a former partner of Andreessen Horowitz, wrote in 2019.
It’s a central paradox for language assistants like Alexa or Google Assistant: the technology has been enormously successful. Many consumers find it really useful. And no one can make money from it.
Amazon can’t get people to buy more stuff with Alexa, and it’s chosen not to require Amazon Echo users to have an Amazon Prime account, the only other credible way to make money. Amazon Skills fizzled out when developers realized there was no money to be made from building the platform. Amazon uses voice queries to help with ad targeting, but advertising accounts for a small percentage of total sales for Amazon.
Amazon is not alone. Google has shifted spending away from its Google Assistant to improve its Google Pixel smartphone, The Information reported earlier this year. Other wannabe Alexa competitors, like Microsoft’s Cortana or Samsung’s Bixby, are essentially dead.
The only company that doesn’t seem to suffer from the voice assistant blues is Apple with Siri. That’s because Apple knows its smart speaker isn’t a loss leader. The original HomePod sold for $349 and the MiniPod for $99. For comparison, you can now buy an Alexa Dot for $15.
Bezos told shareholders in 2013 that Amazon sold its devices to break even because “we want to make money when people use our devices — not when people buy our devices.” Apple took the opposite approach, deciding that if it wanted to sell anything to play music, set timers, and check the weather, it needed to keep margins healthy.
Amazon, hoping to lure users in with a low price, made Echo and Alexa commodities: cheap and unobtrusive.
There is money to be made from selling standard tech products – many companies stay in business selling laptops, home office printers and wireless routers. But as Amazon faces a new era of cost discipline, a “glorified clock radio” can’t lose $10 billion a year or employ 10,000 people.