It’s only four years since Amazon unveiled the Echo, the first wireless speaker to incorporate a voice-activated virtual assistant. Initial reactions were muted, with one press report from the time noting the product’s uniqueness while describing it as “a tad baffling”, but the Echo, and the Alexa virtual assistant it contained, quickly became a phenomenon, with the level of demand catching even Amazon by surprise. So what are conversational platforms in business and how are they developing?
Virtual personal assistants, like Amazon’s Alexa or Google’s Assistant, can schedule appointments, provide weather updates and play music based on voice commands, while a growing number of companies across industries are implementing virtual agents in areas such as customer service and human resources.
Conversational platforms employ a variety of technologies, including speech recognition, natural language processing (NLP), contextual awareness, and machine learning, to enable human-like interactions with computer systems.
Why do conversational platforms matter for business?
Today, nearly every leading technology company is either already producing a smart speaker or developing one, with Facebook the latest to enter the fray, launching its Portal device in October 2018. The appetite for smart speakers is also not limited by geography, with China, in particular, emerging as a major marketplace. Our estimates suggest that the global installed base for smart speakers with hit 100 million early next year, before surpassing the 200 million mark at some point in 2020.
A major factor in the smart speaker’s path to ubiquity has been the willingness of vendors to slash prices, even if it means making a loss on every device sold. The reason for this is simple: the value of smart speakers extends beyond the hardware itself.
Smart speakers give companies like Amazon, Google, Apple, and Alibaba access to a vast amount of highly valuable user data; they allow users to get comfortable interacting with artificial intelligence tools in general, and virtual assistants in particular, increasing the likelihood that they will use them in other situations, such as the car or the office, and they have the potential to lock customers into a company’s ecosystem, making it more likely that they will buy complementary products or access other services, such as online stores.
With the likes of Samsung and Microsoft all expected to launch smart speakers in the next year or so, the market landscape will continue to fluctuate. It is also likely that we will see two markets emerge: on one hand, the cheap, impulse-buy end of the market, used by vendors to boost their ecosystems and, on the other, the more expensive, luxury end of the market, where more focus is placed on sound quality and aesthetics. This is the area of the market at which Apple has aimed the HomePod and early indications are that this is where Samsung’s Galaxy Home will also look to make an impact.
What are the big themes around conversational platforms?
The battle for virtual assistant supremacy
In January 2018 it was reported that both Amazon and Google had cut the prices of their cheapest smart speakers, the Echo Dot and Home Mini respectively, to such an extent over the Christmas period that they were actually losing money on every device sold. In China, Alibaba slashed the price of its Tmall Genie X1 by 80% to 99 yuan, about $15, during the country’s Singles Day shopping event. This clearly illustrated that the smart speaker strategy for these companies is less about making money from product sales, and more about building markets for their virtual assistants. Smart speakers, particularly lower-priced models, are gateway devices, in that they allow consumers to become comfortable interacting with a virtual assistant, like Amazon’s Alexa, in a safe environment and thereby increase the likelihood that they will use it in other situations, such as in the car or the office. For tech companies serious about competing in the virtual assistant sector, a smart speaker is becoming a necessity.
It’s all about the ecosystem
For most of the leading players in the market, a smart speaker is just one part of a broader ecosystem of products and services. Once consumers have purchased a device, they are more likely to use the associated ecosystem to carry out tasks such as listening to music, checking their calendars or making online purchases, with much of this activity directly driving revenue for the owners of said ecosystem, as well as giving that company access to huge amounts of extremely valuable consumer data. According to figures from Consumer Intelligence Research Partners (CIRP), the annual spending on Amazon.com of the average US-based Echo owner was approximately $1,700, compared to $1,300 for the non-Echo owning Prime member and $1,000 for the standard Amazon customer. This is another compelling reason for companies like Amazon to accept making a loss on smart speaker sales, as the potential revenue derived from other areas will be more than enough to compensate.
As you would expect, manufacturers design their smart speakers so they perform best within their own ecosystem, often at the expense of other services and products. So, while it is possible to access, for example, Apple Music through a Google Home device, the process is inconvenient and unintuitive whereas, with a HomePod, Apple Music works seamlessly. Likewise, non-Apple Music subscribers who buy a HomePod will be more likely to subscribe to the service, for the sake of convenience if nothing else.
How smart is your speaker?
For a wireless speaker to be classified as “smart”, it needs a virtual assistant, but just how smart are these tools? The good news is that, thanks to recent advances in speech recognition, the leading devices are able to understand the vast majority of what is said to them. The challenge comes when they need to provide a response to questions or queries, and here there is still room for improvement. In a December 2017 test carried out by Loup Ventures on three smart speakers, the Amazon Echo, Google Home, and Harman Kardon Invoke, which incorporates Microsoft’s Cortana assistant, the best-performing device, the Google Home, provided the correct answer 81% of the time. This was well ahead of the Echo and Invoke, which scored 64% and 56% respectively. However, it was notable that the performance of all three assistants had improved since tests earlier in the year, with the total number of correct responses increasing by 29% for the Echo and 41% for the Home. The suggestion is that, as more people interact with their smart speakers, it will help train the virtual assistants and thus improve their performance.
The automated home
Smart speakers were expected to make the concept of the automated home more mainstream, by enabling simple, voice-activated control of products such as lights, locks and thermostats and, thanks to their relatively low price-point, offering an ideal entry point for customers interested in making their home “smarter”. Early indicators are that this prediction will prove to be accurate, a May 2018 study by PwC found that 40% of those in the UK who entered the connected home market did so through smart entertainment devices, including speakers. Another survey of UK consumers, this time carried out by YouGov, found that 34% of smart speaker owners used it to interact with other smart devices. This number is likely to rise in the coming years, with smart speakers increasingly operating as a hub for connected home technology, but concerns about the overall utility of smart devices compared to more traditional alternatives, particularly given the cost disparity, and frustration dealing with technical issues, particularly around installation, will continue to hamper adoption.
Smart speakers, like any internet-connected piece of technology, are vulnerable to cyberattacks. For consumers, security is a major concern, which will either put them off buying a smart speaker entirely or make them more cautious about where and how they use them. Technology vendors, therefore, need to reassure people that their devices are as secure as possible, both by patching existing flaws rapidly and effectively and ensuring that security is prioritized when developing new products. As with other cybersecurity threats, users also have a responsibility to ensure that their devices are running the most recent firmware and that they take due care to avoid accessing malicious versions of legitimate applications.
The language barrier
One of the major challenges for smart speaker manufacturers looking to enter new geographic markets is the time and effort required to train conversational platforms to speak and understand additional languages, including distinguishing regional accents and dialects and handling colloquialisms. Services also need to be localized so that responses to queries are appropriate to the user’s location. For example, if a virtual assistant is asked to spell a word then the correct response could differ depending on whether they are in the US or UK. It should be noted that consumers in regions where local language support is not available can still buy devices from the likes of Amazon, since the company announced in December 2017 that it would ship Echo speakers to 80 new countries as long as customers ordered through its US, UK or German sites. So, for example, an English speaker living abroad would still be able to buy and use an Echo, while local consumers able to speak English as an additional language could also interact with the device.
Differentiating in a crowded market
For companies entering the smart speaker market, differentiating their product from those already available can be difficult. In terms of design, most speakers fit into one of three categories: the cylindrical look, similar to a flask or a tin can, popularized by the original Amazon Echo, the compact hockey puck style adopted by the Echo Dot and Google Mini, and the speaker with a screen, first seen in the Echo Show and recently adopted by Lenovo’s Smart Display device and Google’s Home Hub. Some companies have attempted to push the envelope with regards to the overall look of their speakers, Baidu’s Raven H, for example, was made up of eight differently-coloured squares, stacked one on top of the other, with the top square containing an LED display, while others have tinkered around the edges of the formula, as with the Echo Spot, a compact device featuring a small screen, but overall there has been relatively little innovation when it comes to aesthetics. Some companies have instead chosen to emphasize the audio capabilities of their devices as a way to stand out from the crowd. Sonos, for example, looked to build on the excellent reputation of its range of wireless speakers when it launched the One, a smart speaker with Amazon Alexa built in and plans to add support for Google Assistant, giving users a choice of which assistant to use, while Apple has made sound quality one of the main selling points of its HomePod. The other area where there is scope for differentiation is the price.
What is the history of conversational platforms?
The smart speaker market is still in its infancy, with the first model arriving in 2014, and even then it was only available on a limited basis, but it has grown rapidly, with the installed base expected to pass the 100 million mark early in 2019. The emergence of the smart speaker is tied in with developments in AI, specifically around conversational platforms, which employ a range of technologies, including speech recognition, NLP, contextual awareness and machine learning, to enable human-like interaction with computer systems. For this reason, milestones such as the launch of the first virtual assistant, Apple’s Siri, are included in the timeline for the smart speaker theme, shown below.
The smart speaker story… … how did this theme get here and where is it going?
- 2011: IBM Watson beats human champions in the TV game show Jeopardy.
- 2011: Apple’s natural language-based virtual assistant Siri appears on the iPhone 4S.
- 2012: Google releases Now, a voice-activated assistant that proactively delivers information to users.
- 2013: Microsoft demonstrates the Cortana virtual assistant at its annual developer conference.
- 2014: Amazon launches Echo, its intelligent voice-activated speaker, which include the Alexa virtual assistant.
- 2015: Having previously been available on an invite-only basis, Amazon’s Echo goes on general sale in the US.
- 2016: Amazon expands the Echo line with the Dot, a smaller alternative to the Echo, and the Tap, a portable version.
- 2016: Google Assistant debuts as part of the messaging app, Allo.
- 2016: Google launches the Home, its answer to the Echo. It is released in the US in November 2016.
- 2016: The original Echo becomes available in the UK and Germany.
- 2016: South Korea’s SK Telecom reveals its Nugu virtual assistant and smart speaker.
- 2016: Samsung acquires virtual assistant startup, Viv.
- 2016: The JD.com-owned LingLong launches China’s first smart speaker, the DingDong.
- 2017: KT joins SK Telecom in the South Korean smart speaker market with the Giga Genie.
- 2017: Bixby, Samsung’s virtual assistant, debuts on the company’s Galaxy S8 smartphones and Note 8 devices.
- 2017: Google Home made available in the UK, before heading to Canada, Australia, Japan, France, and Germany.
- 2017: Baidu reveals its first entry into the smart speaker market, the Xiaoyu Zaijia (or Little Fish).
- 2017: Harman Kardon’s Invoke, featuring Microsoft’s Cortana assistant, is launched.
- 2017: The Echo family continues to expand with the launch of the Look, Show, Spot, and Plus.
- 2017: Amazon also introduces the second generation of its original speaker, the Echo.
- 2017: Apple announces the HomePod, with initial releases planned in Australia, the UK and US.
- 2017: South Korea’s Naver releases its first smart speaker, the Wave, featuring its Clova virtual assistant.
- 2017: Alibaba shakes up the Chinese smart speaker market with the Tmall Genie.
- 2017: SK Telecom launches the Nugu Mini, a portable version of its smart speaker.
- 2017: Sonos announces the One, which initially comes packaged with Amazon’s Alexa assistant.
- 2017: Google expands its smart speaker range with two new devices, the Home Mini and the Home Max.
- 2017: Naver adds Friends to its line of Clova-powered smart speakers.
- 2017: Kakao’s Mini AI speaker goes on sale in South Korea.
- 2017: Baidu unveils the Raven H, the most unusual-looking smart speaker to date.
- 2018: Google announces it is working with JBL, Lenovo, LG and Sony on smart speakers with displays.
- 2018: Google also reveals Assistant will be able to speak more than 30 languages by year-end.
- 2018: Apple’s HomePod begins shipping. By mid-year, it is available in six countries.
- 2018: Samsung announces the Galaxy Home, which will feature its Bixby virtual assistant.
- 2018: Huawei enters the smart speaker market with the Alexa-powered AI Cube.
- 2018: Facebook’s Portal products, which incorporate Alexa, go on sale in the US.
- 2019: The smart speaker installed base is expected to hit 100 million.
- 2020: Installed base expected to pass 200 million.
This article was produced in association with GlobalData Thematic research. More details here about how to access in-depth reports and detailed thematic scorecard rankings.