In Australia, though Google Assistant dominates with over half the market, Amazon’s Alexa comes in second.
Apps utilising the Alexa Assistant are downloaded by the millions and device sales are steady, if not growing.
In the US, over forty percent of people use a smart speaker, with over 40 million Alexa Voice Assistant users – that’s about 10 percent of the population.
That kind of market dominance is something Amazon is used to – and yet virtual assistants haven’t delivered on profits.
And now, in the post-COVID big tech market retraction, Alexa is gutted.
Over 18,000 people will be laid off from Amazon in the coming year with the areas servicing Alexa firmly on the list.
Alexa-related devices – which have been historically sold at cost – will have to show a profit like other sectors of the behemoth.
Just a few short years ago, the digital voice assistant was all the rage, with “Hey Google!” competing with “Alexa…” all over major trade shows like CES Las Vegas.
Alexa came in everything from appliances to cars to, well, Amazon’s own Echo devices.
Google’s level of market dominance is expected, as Android phones coupled with global search brand familiarity makes its name a literal word in almost every language (indeed appearing in dictionaries).
Amazon also did very well for itself in the US, leveraging its retail market dominance to gain traction domestically.
Nonetheless, COVID should have made us cosier with our voice assistants.
The number of users did grow, and a number of Alexa-affiliated appliances did sell.
So why did Amazon lose billions of dollars a year on the black box with the soft voice?
Why Bezos lost his bet on Alexa
Jeff Bezos predicted that Alexa would ease her way into people’s homes as a device that could tell you the weather, play music, set a timer or get basic information from the internet with just your voice.
And it did that – well enough for those things to be the primary reason people use Alexa.
The key money-making features – shopping and home automation – never sufficiently reduced purchasing friction to make a difference in sales.
The lack of orders could partly be because of the primary shoppers in the home.
They also were less likely to shop by voice.
Complex ordering, selecting the correct items, and price comparisons is harder to get right without an interactive interface.
The best it gets is “order again” and what happens if the toilet paper is $62?
Finally, in Australia anyway, the supply chains and costs didn’t make ordering household items worth it.
Unlike in the US, where you could literally get something like toothpaste within hours of ordering it, in Australia you might be waiting days for deliveries to reach you.
More localised competing services, such as Coles delivery, outdid Amazon easily.
In response to this, Alexa has spent a couple of years using a harder upsell in some markets.
Ask about the weather and she might respond with a suggested purchase from Amazon.
It is a constant reminder of one key fact: if you think Alexa works for you, you’re delusional.
She’s your roommate who works for Amazon and is always trying to close a deal.
“Alexa…turn off all lights…Alexa…? ALEXA!”
So, what will these layoffs mean to the average Alexa user?
Don’t expect Alexa to understand you better.
Amazon’s NLU (Natural Language Understanding) staff is going to be slashed.
So, if you need to imitate an American accent to be understood – well, you might have to continue to do that.
While Alexa’s abilities to understand customers has arguably been degrading over time, Google is introducing new features in 2023, such as pre-set phrases (for which you do not have to preface with “Hey Google”) and better understanding of pauses in speech.
Never mind that that many of the “fundamentals are still broken”, says Tigger Kindel, one of Amazon’s own.
So, in the short term, Alexa won’t be getting better at understanding your Aussie lingo; but it might not understand your Chris Hemsworth American accent either.
Don’t take it personally.
“Alexa, remind me when Amazon devices go on sale.”
Amazon Alexa hardware prices will go up, but third-party devices are likely to be introduced.
Amazon can’t afford to let Siri or Google gain more ground than they already have.
It is likely to claw back some profit by raising the prices of Amazon-manufactured devices.
Making money on the hardware will be some compensation toward its big investment under the assumption that consumers will continue to buy the devices after sales at cost have set expectations so low.
Regardless, it is unlikely to fill the Alexa hole, tens of billions of dollars deep.
“Alexa, who’s listening to this conversation?”
Alexa is unlikely to get more privacy-conscious or cyber-safe – and it might become less so.
The current approach sees Alexa “skills” isolated from other skills and the rest of the Amazon ecosystem.
So, when you order groceries, it is isolating that interaction and is using privacy and safety protocols in accordance with Amazon’s terms and conditions.
However, with third party vendors increasingly in the mix, “Alexa” isn’t one person anymore, it’s a crowd – and you might not know exactly who you are talking to anymore.
Your habits or information may be more vulnerable than before.
Privacy and cyber-conscious individuals would eschew the Alexa, or any virtual assistant.
Most will shut Siri off and look out of the window to check the weather or press a button to turn a light on and off.
But for those who don’t, there might be no telling who is listening – or when – to your special little box with the soothing voice.
Nathalie Collins is an adjunct Senior Lecturer at Edith Cowan University’s top-ranked School of Business and Law. She has consulted to industries as varied as biotechnology, Healthcare IT, Tourism and Consumer retail.
Jeff Volkheimer is the Senior Director Collaboration and Continuity Technologies at Duke Health in North Carolina.
Paul Haskell-Dowland is the Professor of Cyber Security Practice at Edith Cowan University in Perth, Australia; and a Senior Member (CP) of the ACS.