Alexa, Google: everybody now has their favourite, but why are people drawn to one or the other? And which should you invest in?
Amazon is currently dominating the voice-controlled smart speaker market, expecting to end the year with a 70.6% share of the 35.6 million-device US market to Google’s 23.8%. The game is currently Amazon’s to lose, but after having the market all to itself for two years it’s now facing competition from the other big tech giants. What’s keeping them ahead, and how long can they hold on to their lead?
“Amazon has a 70.6% share of the smart speaker market to Google’s 23.8%. ”
Amazon has the short term advantage
Most market watchers expect Amazon to continue leading for the next few years at least. It currently enjoys first-mover advantage, having launched its speaker two full years ahead of Google; it enjoys widespread customer awareness; has formidable retailing power and a clear commercial strategy - Echo owners spend 10% more on Amazon after buying the device.
“Echo owners spend 10% more on Amazon after buying the device. ”
Google has long term advantages
Google is playing catch-up. Lagging in customer awareness, market share, and retailing power, it is relying on its technology to give it the edge as the market matures. Google has long been investing in AI and natural language understanding technologies - as the world’s largest search engine Google's business is figuring out what people want and giving it to them - and this technological heft will help it in the long run. What’s less clear is what commercial role voice plays at Google. Google makes its money in advertising - and the prospect of ads on voice devices is not going down well with consumers. Where Amazon has clear commercial sense for the Echo, Home is leaving some investors scratching their heads.
“As the world’s largest search engine, Google’s business is figuring out what people want.”
How do the device experiences compare?
While the strengths and weaknesses of the two companies will be major factors in determining who leads the market, success will come down to the experiences offered by the devices themselves.
There are thousands of articles comparing the capabilities, prices, and skills of the Google Home and Amazon Echo, but what has been interesting to us is the assistants themselves: their voice, tone, mannerisms, and feel. Since the Google Home joined the Amazon Echo in our studios, both have been handling a daily stream of requests: from the mundane, to the challenging, to the bizarre. Everybody now has their favourite, but why are people drawn to one or the other?
Wakewords - Alexa wins
Alexa is called to attention with the wakeword ‘Alexa’, while ‘Ok Google’, or ‘Hey Google’ brings the Google Assistant to life. From the off, waking Alexa feels much more natural. ‘Alexa’ is not only shorter and snappier, it also feels more like you are talking to a real person. In comparison, ‘Ok Google’ feels awkward and can feel confrontational. ‘Hey Google’ feels much more natural, yet still a little cumbersome when asking multiple questions one after another.
“Saying ‘Alexa’ is not only shorter and snappier, it also feels more like you are talking to a real person.”
Natural speech - Google wins
I first started using Alexa when I bought an Amazon Echo for my own home, and was immediately surprised and pleased by how human the voice sounded. Sat for a few hours next to the Google Assistant, however, and Alexa begins to sound a little robotic and monotonous. Google’s voice rises and falls in tone, the pace of its responses is less faltering, and its phrasing is a little more humanised than Alexa’s, which begins to sound a little stiff and formal.
Google does, however, speak quite rapidly. While this is appreciated in certain situations, when it lists things or reads out longer pieces of information it becomes harder to keep up with. Alexa trips up a little with lists too, but generally speaks at a more sedate pace, giving a bit more time to digest its responses.
Physical device feedback - Amazon wins
Although I’m focusing on the voices of the two assistants, I want to throw in a quick note on the devices themselves. When woken, Google Home lights up with four coloured lights, which spin when it’s ‘thinking’. Echo illuminates a blue ring of lights, with the light closest to you lit the brightest, and the full light ring has a ‘dappling’ effect when it is thinking of an answer. The differences may seem trivial, but the more I used them, the more Echo gave me the impression it was paying me attention more than Home. Home was listening, while Echo was listening to me. Home’s thinking animation is too reminiscent of a loading icon on a computer screen, while someone in our studio summed up Echo’s animation as ‘a visual hmm’. For us this gives Echo the edge in humanising the product’s visual interface.
“Home was listening, while Echo was listening to me.”
Reassurance of understanding - Amazon wins
Given the scope for misunderstanding, it’s useful for the assistants to have a way of confirming they heard the right thing without having to refer to the accompanying app. Of the two, Alexa handles this better, while Google is hit and miss (albeit with more hits than misses). For example, when asked ‘What is the capital of the United Kingdom?’ both rephrased the question back to me in their answers — ‘London is the capital of the United Kingdom’. This gives me complete certainty that they understood my words and intent — if they had just answered ‘London’ there is no way of knowing what they heard, and therefore knowing the answer is correct.
On a related note, neither assistant could tell me what I just asked it. Variations of ‘What was the last question I asked?’ drew a blank from Alexa, while Google started telling me about The Last Question by Isaac Asimov.
Complexity of requests - Amazon wins (just)
Neither assistant is very satisfactory at handling more complex questions. ‘What are the populations of France and Italy?’ stumped both. A couple of times Alexa managed an answer; ‘What is the population and area of France?’ gives a correct answer, as does ‘What is the capital of the country Glasgow is in?’, Google either doesn’t understand the question or begins reading out irrelevant information from third parties.
Completeness of answers - Both win
When looking for the completeness of answers, Alexa does well on the basics, while Google outperforms on more off-the-beaten-track questions. Where Google provides answers from third parties it adds in extra information, which isn’t unwelcome, but in other cases it gives only a single answer, when I know there were multiple.
“Alexa does well on the basics, while Google outperforms on more off-the-beaten-track questions.”
For example, when asking it to define certain words Google offers up only a single definition. Further prompting and rephrasing of the question manages to get it to refer to Dictionary.com where it gives more definitions, but Alexa automatically rhymes off all the definitions it can find, which was the experience I was expecting.
Conversational flow - Google wins
One major advantage of the Google Assistant over Alexa is that it can string a basic conversation together. When asked ‘Who is Donald Trump?’ it gives accurate information from Wikipedia, then when asked the follow up question ‘Who is his daughter?’, is understands that I am still talking about Trump, and names his two daughters. I tried a bunch of these questions with follow-ups and it can handle most, then I came across a quirk. ‘What is the French word for autonomous?’, gives a correct answer, but “How do you spell that?’ gives ‘That is spelled t-h-a-t’. If Google is going for dad jokes they nailed it.
“One major advantage of the Google Assistant over Alexa is that it can string a conversation together.”
The only awkwardness in these conversational threads is the need to keep using the wakeword before each follow-up question, which breaks the back-and-forth flow.
Accuracy of answers - Both win
Although the Google Assistant can answer a much wider range of questions, I found Alexa to be more accurate in its answers. While Alexa appeared to be drawing its information from specific databases, Google was pulling in info from around the web, some of which was outdated, inaccurate, or off-topic.
Failure responses - Google wins
Conversational interfaces such as Alexa and the Google Assistant can be asked anything, even nonsense questions, and users will always expect a response even when the assistant can’t understand them. ‘Failure’ responses can be tricky to get right — it might be easy to craft an amusing or whimsical answer, but when the user has heard that answer ten times in a row it rapidly loses its novelty.
“It might be easy to craft an amusing or whimsical answer, but when the user has heard that answer ten times in a row it rapidly loses its novelty.”
Both (thankfully) avoid cutesy answers in these situations, but Google’s are a little more dynamic. Alexa repeats statements saying it can’t understand; Google is a little more optimistic, caveating its lack of understanding by saying that it’s learning all the time and hopes it would be able to answer it soon.
In all, Google’s failure responses are better, it gives off a much stronger impression of its learning abilities and sounds more human. Next to it, Alexa’s answers begin to sound a little ‘computer says no’.
Final Impressions
Overall, Google edges Amazon in terms of user fidelity - Home and the Google Assistant powering it are just a bit easier and more natural to talk to, but Amazon’s Echo/Alexa combo scores some big wins and has a strong lead in the market.
“Google kept disappointing me, but in a good way. ”
Google kept disappointing me, but in a good way — the more I used it the higher my expectations of it became, whereas after a while of chatting I had a lower bar for Alexa so it was more likely to surprise me.
Both assistants are in their early days, and Amazon and Google are improving them at such a rate that any real review of them is likely to be outdated not long after publication. Both have their pros and cons, and both Amazon and Google have weaknesses and strengths which will help shape the future development of voice interaction, not to mention the recent arrival of Apple and Microsoft to the market. For anyone working in the voice interaction space, both assistants offer examples of good practice and avenues for improvement, but for such a young industry they deliver an optimistic start.
Which should you invest in?
While Google delivers a better user experience, Amazon has a larger ecosystem behind its assistant. If you're buying one for your home (and we recommend that you do!) consider what your motivations are, if you are looking for a device to chat to and learn some interesting trivia then opt for the Google Home. If you are looking for a functional device to help plan and manage your smart home, Amazon Echo is the way to go.
For companies, investing in Amazon Echo is an obvious choice - it has the market share and is quickly gaining some pop culture credentials - but Google are catching up quick and in a market in its disruptive early days, there is everything to play for.
Screenmedia is an innovation, design, and development practice. If you are looking to explore new digital media including voice, wearables, and the Internet of Things, get in touch to see how we can support you.
Andrew
Innovation Lead