Image source @ Visual China
Wen 丨 Wang Xinxi
Since around 2016, smart speakers have developed for 5 to 6 years.
In 2014, Amazon's smart speaker Echo swept the world, attracting the attention of domestic Internet giants. Since the second half of 2016, a small number of players have begun to test the waters into this battlefield, and in 2017, mainstream Internet manufacturers have gone down.
According to statistics, this year, there are as many as 50 domestic smart speaker companies.
From 2018 to 2019, smart speakers entered a climax of development, according to the report of market research institute Canalys at the time, in the second quarter of 2019, China's smart speaker market performed better than other countries, and shipments were more than twice that of the US market.
According to the data of Aowei Cloud Network, the retail sales volume of China's smart speaker market has increased rapidly from 1.76 million units in 2017 to 37.7 million units in 2020. In 2017, China's smart speaker retail sales were 490 million yuan, and by 2020, the retail sales of smart speakers have exceeded 7 billion yuan.
Today, the domestic smart speakers generally show a three-legged trend. According to the data of the China Commercial Industry Research Institute, in 2020, the market share of Alibaba Baidu Xiaomi was 34.96%, 33.4% and 27.2% respectively.
Despite the rapid growth in retail sales, from 2020, the popularity and attention of smart speakers, both domestic and global, have shown a downward trend.
The data shows that in 2019, global smart speaker shipments set a new record of 146.9 million units, and global smart speaker sales exceeded 150 million units in 2020. Growth has slowed markedly.
Domestic data also show this tendency. According to IDC data, after 2020, the growth of the domestic smart speaker market began to show signs of decline. In 2019, domestic smart speakers also showed triple-digit growth, reaching 109.7%. By 2020, the overall domestic smart speakers have dropped by 8.6% compared with 2019.
Even from Amazon, the leader of smart speakers, its retention has begun to have problems.
According to Bloomberg, Amazon's internal analysis of its intelligent voice service Alexa business shows that the business has recently been less profitable, the user churn rate is high, and the intelligent voice market will grow by only 1.2% in the next few years.
What we can clearly feel is that smart speakers are not hot, and there is even a tendency to go downhill.
Why aren't smart speakers on fire?
In fact, we have observed that 5 to 6 years have passed, and the development of smart speakers has now, its transformative and innovative, product experience has not actually met the user's expectations for such products.
In today's domestic market, smart speakers have made some progress from content entertainment services to smart homes, voice interactions, etc., but there are no innovative products that exceed expectations.
Behind the rapid growth of smart speakers, it is closely related to subsidies and price wars.
The data shows that in the first half of 2019, the average sales of a single device of smart speakers was 193 yuan. Today, from Xiaomi Xiaoai smart speakers to small smart speakers and Tmall genie smart speakers, the pricing is generally between 95 and 499.
Today, small factories that can't afford to play the price war have basically left the market, and large factories are generally selling hardware at a loss. However, whether it is brand height or product influence and functional innovation, it is not yet successful.
What are users' expectations for smart speaker products?
From the initial imagination of people's smart speakers, it will become the core entrance of smart homes, through smart speakers, can control all electrical equipment in the home.
Of course, if this goal is too ambitious, then in the user's expectations, the smart speaker can at least have a smooth dialogue with the room owner through voice interaction and voice recognition of the room owner's instructions in different locations. But this small goal, smart speakers also do not do well.
Today, smart speakers have not iterated on the voice interaction experience.
Judging from the ranking of smart speaker use functions in foreign voicebot.ai in 2019, users still have not left routine functions such as listening to music, asking questions, checking the weather, setting alarm clocks, and listening to radio when using smart speakers.
What are the problems with smart speaker products on the market now?
First of all, it needs to be admitted that the current smart speaker is still immature at the level of voice interaction, although the popularization of the speaker with the screen, but the integration of voice and buttons and touch screen is not good, weakening the characteristics and evolution of voice interaction.
Some industry practitioners pointed out that the current smart speakers almost do not support complex sentences, negative sentences, omitted sentences, multi-round dialogue and other natural language input, and do not play the real advantages of voice interaction.
This is probably the common feeling of consumers who have used smart speakers - whether it is long sentence recognition or semantic understanding, answering questions is a high probability event, and the experience is not good.
The voice interaction and recognition rate are not good, and the user's natural interest in using it is not great.
According to a data released by an industry practitioner, although the number of smart speakers in the world exceeds 600 million, there are only about 10 interactions per day. This is the global data, in the overseas smart speaker market, is the Amazon Echo, Google Google Home, Apple HomePod dominated by the market, in the domestic, this data is probably lower.
Speech interaction and long sentence recognition rate have not been improved, and it is also related to the lack of massive data training.
For Internet manufacturers, smart speakers are an independent product, can not use the massive data of the mobile Internet, it needs to respond to the knowledge graph data of questions and answers for training.
If the recognition rate of smart speakers is low at the beginning, user stickiness is low, and user interaction frequency is low, naturally there is a lack of massive data to iterate on its intelligence.
At the intelligent level of voice interaction experience, Amazon has not brought about a fundamental voice experience change, such as its inability to recognize people or the voice and needs of specific individuals, and its inability to have a payment experience comparable to that of smartphones.
In the eyes of many consumers, Alexa's practicality is insufficient.
According to Bloomberg Businessweek, for years, most Alexa users used only voice devices to play music, or set a timer while cooking, or turn on the lights.
In a 2019 planning document, Amazon employees noted that new Alexa users will find that within three hours of activating their devices, they are using only half the full functionality.
Today, intelligent voice assistants are the standard for mobile phone products, and they are almost standard for in-vehicle intelligent systems.
However, from the experience point of view, smart speaker voice interaction only stays at the basic level of conventional voice command recognition, the current smart speaker can not take the initiative to serve, can only be passive operation, the degree of intelligence is not high, the experience is not sexy, and can not see too much potential for innovation and change.
Many kinds of things, resulting in smart speaker products and users expected experience and its transformative significance is far from the current smart speaker more reduced to a dispensable voice + screen small toy.
Smart speakers with screens have become mainstream, what should players reflect on?
Today, from the Amazon Echo Show in overseas markets, Google Nest Hub, to the domestic Millet, etc., various manufacturers are positioning smart speakers with screens as a product for a home entertainment experience. Smart speakers with screens are also considered to be the next stage of smart speakers.
But we should also see that the use of the screen is actually irrelevant to the user's just need for smart speakers.
And the big manufacturers, including Amazon, are dead with screen smart speakers, in fact, its strategy lies in two aspects.
First, in order to make up for the relatively inefficient voice feedback, the screen allows the smart speaker to assume a variety of functions such as audio-visual entertainment, video communication, home monitoring, etc., hoping to cover up the shortcomings in voice interaction through the screen touch experience of this universal product play method that has been verified on PCs, mobile phones, and tablets.
The introduction of the screen is very consistent with the direction of the content ecology of the major giants - whether it is Jingdong Ali, Xiaomi Baidu and other players, their content ecological construction is basically related to "seeing", and less related to "voice interaction".
If it is the main voice interaction, then the content ecology of the major giants on the mobile Internet cannot be well connected. With the screen, the layout of the content ecology can be migrated to the smart speaker product.
For example, Xiaomi Xiaoai speaker integrates video graphic content such as headline video, iQiyi, Youku, and Station B in terms of content;
Xiaodu also has a video content ecology including iQiyi, Good-looking Video, Youku, B Station, Mango TV, Migu, etc.
Tmall Genie also fully accesses Alibaba's content and service capabilities such as entertainment, health, education, and shopping.
Today, smart speakers generally cover the content ecology of communication, music, family interactive games, children's early education, etc., and the introduction of these content levels is also closely related to the revenue model of big manufacturers such as content paid memberships.
But it also makes the smart speaker more and more like a tablet, after all, playing games on the screen, watching videos, swiping graphics, just need a mobile phone or iPad, rather than a voice device.
At the same time, the evolution of smart speakers in voice interaction has slowed down, and the screen has also made the positioning of the product confusing - smart speakers look no different from tablets.
Harvard Business Review once argued that over time, the value of the technology behind smart speakers will be less and less about the device itself, and more about how to let consumers speak directly to any device.
Smart speakers have been considered the bearers of AI technology in the past, but today, after the evolution from screenless to screen-based smart speakers, manufacturers have superimposed video, entertainment, learning, call and other functions, making the product boundaries of screen speakers continue to expand.
But this approach undoubtedly ignores the core battlefield of smart speakers in the voice interaction battlefield, because it is in line with the lazy and natural nature of human beings to complete what they want to accomplish through voice than touching the screen with their fingers.
After the launch of the Echo, Amazon has shown people the great potential of its intelligent voice system, alexa platform, in the past.
Because it is a new way of playing different from the visual touch screen mode, Alexa creates an auditory content ecology, which is actually a voice interaction operating system similar to the Android version.
Therefore, creating a highly intelligent voice system ecosystem should actually be a unique differentiating path for Amazon.
However, as smart speakers with screens become the mainstream, the touch screen control and visual viewing experience dominate, depriving the voice interaction of time and evolving in a more in-depth intelligent direction.
In fact, Amazon has just given up its own differentiated characteristics and core competitiveness, which may be an important factor in the gradual loss of Amazon users.
Judging from the current evolution trend of smart speakers, smart speakers are increasingly inclined to become a cheap version of tablets, video playback has become a core function, and this is also an important battlefield for Apple to catch up with Amazon and Google in the future.
For manufacturers, this can indeed rob some of the users of the tablet market, but where is the difference value between it and the tablet, which is undoubtedly worth the manufacturers to think about.