Once a near non-existent presence in the smart speaker market, with a market share of less than 1% in Q2 2018, Baidu rose astronomically over the year that followed to become the second-largest vendor of smart speakers globally with a market share of 17.3%, behind only Amazon at 25.4%.

In its home market of China, Baidu was for a time the number one smart speaker vendor, surpassing rivals Alibaba and Xiaomi, who collectively accounted for 94% of smart speaker sales in China in the first quarter of 2018. And while Alibaba has since – narrowly – regained its position as market leader, Baidu remains a fierce contender.

Baidu achieved this remarkable feat the same way that rivals like Alibaba and Xiaomi, and overseas counterparts like Amazon and Google, dominated the market and sold millions of smart speaker units: through deep discounts. China’s smart speaker market has long been characterised by a “price war” between its major players, who vie with each other to lower the price of devices and release affordable models that consumers will snap up. Baidu, Alibaba and Xiaomi have been able to keep their prices low by subsidising the cost of smart speakers, trading in revenue in the short term for the ability to influence consumer habits in the long term.

But now, just a few months after it captured a huge chunk of the smart speaker market, Baidu is changing up its strategy and pulling out of the price war, launching its latest smart speaker – the Xiaodu Zaijia X8 – at full price in mid-December. The search giant believes that it can make smart speakers commercially viable – something which none of the other major vendors in China or the west have yet managed to do.

How is Baidu planning to accomplish this, and can it succeed?

The stickiness of smart speakers in China

Smart speakers in both the west and the east have proven to be popular consumer devices, with tens of millions of devices sold every quarter. But in China in particular, smart speakers have proven to be uniquely ‘sticky’ – achieving not just widespread consumer adoption but a high level of repeated usage and engagement.

According to figures cited by Jing Kun, Baidu’s recently-appointed VP of Smart Living, in an interview with the Financial Times, consumers are using their smart speakers an average of 20 to 30 times per day and engaging with them for around two hours in total. This high level of engagement can partly be attributed, according to Jing, to consumers in smaller cities in China, who tend to treat their devices as a cheaper alternative to a television or a sound system. However, voice-activated devices and voice controls in general have always found a natural home in China, as they are better-suited to the Chinese language than keyboard-based inputs, and produce a much lower error rate.

These figures – even if they are high estimates – stand in stark contrast to the dismal levels of engagement reported by the most recent US Consumer Adoption Report from Voicebot, released in March 2019, which found that less than half (47.4%) of smart speaker owners in the United States use their devices even once per day; the remainder use their devices either monthly (26.1%) or “never or rarely” (26.5%). This shows that, while the low price point might have been successful in driving sales for smart speaker vendors like Amazon and Google, western consumers aren’t finding many reasons to integrate speakers into their everyday lives.

In his October interview with the Financial Times, Jing was already asserting that smart speakers in China had “crossed the chasm” to becoming a commercially viable product. Now Baidu is ready to put its money where its mouth is (so to speak), confident that it has successfully proven the value of its devices to the extent that it can stop subsidising them so heavily.

While some experts, such as Beijing-based researcher Sophie Pan, have cautioned that the market is still in a phase of “educating” consumers and cultivating habits, Jing told China-based publisher 36kr that “the period to educate consumers with price wars is over”, adding that it is “impossible for us to continue to subsidise [smart speakers]”.

Other markets in China that were heavily propped up by subsidies, such as the bike-sharing market, landed companies with crippling debts when the bubbles burst, something that Baidu is clearly keen to avoid. Despite its recent smart speaker success, the company is in a precarious position at the moment: a changing internet landscape that has seen consumers rely much less on search engines for information recently led to Baidu posting a loss for the first time in 14 years, and it has seen its ad revenues slide behind up-and-coming TikTok owner ByteDance. For this reason, it can’t afford to keep absorbing losses from its smart speakers in the same way that companies like Alibaba and Xiaomi can.

But Baidu’s strategy for smart speaker profitability doesn’t just rest on hoping that consumers will be willing to fork out more money for its devices.

Smart displays, subscriptions and entertainment

China’s major smart speaker vendors are each focusing on carving out a different niche for themselves that they believe will be profitable going forward. Alibaba, true to form as China’s largest ecommerce player, has focused on integrating ecommerce functionality into its smart speakers, allowing consumers to connect their Alipay accounts and shop online via voice. Electronics company Xiaomi has focused on promoting its speakers as a smart hub that can connect to a range of smart appliances, such as fridges, air purifiers and vacuums.

Baidu, meanwhile, is marketing its smart speakers primarily as entertainment devices, promoting its range of speakers with in-built smart displays and emphasising their child safety features (such as a Children’s Mode that will block certain content and limit screen time) and suitability for the whole family. Figures analysed by Canalys show just how much emphasis Baidu has placed on smart displays compared with its rivals, both at home and abroad: 61% of Baidu’s smart speaker product mix feature smart displays, nearly three times higher than the worldwide average of 22%. Amazon, Google and Xiaomi all hover around this average at 21%, 19% and 18% respectively. Clearly, Baidu is committed to smart displays as the key to its smart speaker strategy.

Source: Canalys Newsroom

In February, it even released the Xiaodu TV Mate, a three-in-one AI-based home theatre system that combines the visual display of a high-definition TV set with the voice control functionality of a smart speaker.

For video content, Baidu is drawing heavily on its ownership of iQiyi, China’s second-largest video streaming platform, which has continued to be a strong component of Baidu’s business and drove much of its revenue growth in the second quarter of last year. Notably, the latest version of DuerOS, Baidu’s voice platform, offers enhanced voice-controlled playback of iQiyi video content on select devices.

Now that Baidu is no longer subsidising its smart speakers, Jing Kun has indicated that Baidu will begin offering free membership of iQiyi to consumers who buy a Xiaodu smart speaker. This is similar to moves made by both Amazon and Google last year to offer free membership of Amazon Music and YouTube Music, respectively, with the purchase of an Amazon Echo or Google Home.

Baidu also offers music streaming, online games, and educational content, and Jing revealed in his interview with the Financial Times that Baidu plans to experiment with monetising subscriptions in these areas as it evolves its approach to commercialising smart speakers.

DuerOS and conversational excellence

Now that Baidu can no longer rely on pricing as an incentive for consumers to purchase its smart speakers, it needs to be able to show that it can deliver an unparalleled experience that none of its competitors can match. Which is why Baidu has been progressively building up the capabilities of its DuerOS voice platform, encouraging developers to build integrations with the platform, and integrating DuerOS into a wide range of third-party devices.

When Baidu unveiled DuerOS 5.0, the latest version of DuerOS, last July one of the headline features was its ‘full-duplex’ mode, which allows the assistant to detect additional queries or follow-up commands and respond without the need for a wake word, greatly enhancing its conversational capabilities. Other advanced capabilities include the ability to recognise ‘in sync’ when to act on a task, and when to listen and refrain from reacting.

Alongside the launch, Baidu also announced that DuerOS is now installed on more than 400 billion devices – up from 150 million the previous November – and processes more than 3.6 billion voice queries per month.

Baidu’s background as a search company gives it a major advantage when it comes to building a voice assistant with superior natural language processing and conversational capabilities, as it has a huge, ready-made dataset of search queries to draw on and analyse. And more than anything, this – the experience that Baidu can deliver with its voice controls – may be what determines its success or failure in the smart speaker market. The best technology and content offering in the world can’t substitute for a poor experience, particularly when Baidu’s rivals are already offering compelling and well-established alternatives.

If Xiaomi and Alibaba decide to follow Baidu’s move to withdraw from the price war and rely on other means to differentiate their smart speaker offering, it seems likely that user experience, as well as how well each device can integrate with an ecosystem of smart products that enhance Chinese consumers’ day-to-day lives, will become the deciding factors for success in the voice-controlled market.

Explore Econsultancy’s resources on search marketing and voice