Amazon Devices Did Not Fail Due to DSI, It Was a Platform Bet with a GTM Strategy That Didn't Adapt
Half the devices from this Amazon Fall 2021 family picture have been cancelled or scaled back
In today’s print edition, The Wall Street Journal has a well-written, well-sourced article on Amazon’s $25 billion+ loss in devices over the years (online/paywall link: https://www.wsj.com/tech/amazon-alexa-devices-echo-losses-strategy-25f2581a). While it points out that Kindle and Ring are profitable, it blames the division's lack of profitability on the amorphous "DSI" (Downstream Impact) that buying an Echo might have on other Amazon purchases.
What the Journal misses is that there was another, more strategic rationale for investing so heavily in voice-driven AI: it was a bet on computing platform ownership. If voice assistants had turned out to be a significant stand-alone computing platform, Amazon would be able to control and monetize that investment many times over with an app store rent model, advertising, subscriptions, and direct product sales. It's a similar reason Meta is investing so heavily in XR, as I wrote about here: https://www.techsponential.com/reports/meta. The voice-driven home computing platform opportunity didn't pan out; Amazon established leadership and built third party integrations but there's no Alexa App Store and voice is a poor interface for product purchases. It wasn't a foolish idea; Amazon could certainly afford to make the bet, just as it bet that building cloud architecture might be something that Amazon could rent out, not just use internally. That bet paid off handsomely.
Once you're making a voice computing platform bet, you need to invest in the technology and get it to market. I would argue that the go to market was where Amazon invested for too long. Kindles are profitable both because they are sold for a profit and they are vending machines tied to Amazon's own e-bookstore. However, many of Amazon's other devices are sold at cost or even at a loss, with no direct revenue ties. DSI isn't imaginary; if consumers buy an Echo and use it regularly, they are more likely to use Amazon's digital services that are part of Amazon Prime, and Amazon Prime users spend a lot more annually shopping on Amazon than non-Prime subscribers. There's lifetime customer value there, and it can be quantified. The Wall Street Journal suggests that Amazon over-estimated and double-counted the amount, but it doesn't matter: the DSI was never going to be enough to offset the R&D, operational, and manufacturing costs if the product is sold without a profit margin to start out with.
If an Echo Dot is profitable at $99, it doesn't matter if people only use it to set free timers and ask Alexa to tell them jokes. Instead, Amazon has consistently sold that Dot for $19.99 on Prime Day in order to pull sales from the Google Nest Mini. That pricing strategy makes perfect sense when you're trying to establish a moat around your new computing platform -- and it worked! There are hundreds of millions of Alexa-enabled devices installed. However, once it became clear that it was impossible to directly monetize that platform, Amazon should have shifted to selling its hardware at prices that more than covered its costs. If that meant an end to growth and a smaller installed base, so be it.
The new plan is apparently to see if all those Echos can be monetized directly after all. "Remarkable Alexa," is reportedly going to be a subscription-based service with smarter, generative-AI Alexa and more home automation control. We'll have to see what the offering is and the price point, but unless consumers sign up in droves, Amazon is going to have to stop selling its hardware at a loss and see how much people are willing to pay for free timers, weather reports, radio, and companionship. The other option is that Amazon could insert ads into every interaction, weather the inevitable consumer backlash, and see how many people keep their Echos plugged in anyway.
To discuss the implications of this report on your business, product, or investment strategies, contact Techsponential at avitechsponential.com.
Additional Resources: