China is often referred to as the Wild West. That’s especially true for foreign companies that enter with optimism, only to perish in a market that tends to be both opaque and ruthless. Chinese local startups face that same challenge.
In a WeChat post, the startup described how it had worked with Xiaomi for three months in an undisclosed partnership, which involved the integration of its speech recognition technology into the latter’s smart TV, as well as “combined PR.” During those three months, no formal contract was signed – it was only under verbal agreement.
When Xiaomi finally sent a formal document to Mobvoi, there was almost no time to make any changes before the TV’s press conference.
During those three months, no formal agreement was signed.
“It wasn’t until yesterday, three hours before Xiaomi’s smart TV press conference, that we finally received the agreement,” stated Li Zhifei, Mobvoi’s CEO, in a separate post. “The agreement required us to offer free technology and service for three years, with the rights and interests of our brand name to be discussed separately.”
“If we didn’t sign, we would lose the brand exposure that we were previously promised,” he wrote.
Mobvoi didn’t sign the contract. Three hours later, Xiaomi demoed its latest smart TV, showing off its voice control ability. Audience members applauded as commands to rewind, play movies, and identify celebrities executed without error. There was no mention of Mobvoi.
The Beijing-based startup also claims that previous versions of Xiaomi’s press conference presentation, which displayed Mobvoi’s logo, had been changed. In addition, data from the startup’s backend, which is linked to its speech recognition software development kit, showed that Xiaomi used its technology for its demo.
“Mi TV assumed that Mobvoi was going to stop offering their speech recognition service, since they expressed that they didn’t want to sign the agreement and wouldn’t participate in the press conference,” responded Xiaomi’s TV team in a statement.
According to the Chinese smartphone-maker, Mobvoi wasn’t the only one offering its speech recognition technology to Xiaomi – so was Sogou, a Chinese tech firm known for its Chinese input software. “If Mobvoi stopped its service, all of the speech recognition results would go to Sogou,” wrote the Mi TV team.
“Mobvoi unilaterally exposed the content of a confidential agreement. Mi TV feels confusion and regret for the violation of this commercial agreement.”
Sign before you start
Though China’s startup environment is sometimes messy, starting to collaborate before signing an agreement isn’t typical. What happened between Mobvoi and Xiaomi is even more exceptional, especially given Xiaomi’s well-known and international brand.
“It happens, but it’s not that common,” Ken Xu, managing partner at Gobi Ventures, tells Tech in Asia over WeChat. “It depends on how pressed for time both parties are.” Patent infringements by Chinese companies, on the other hand, happen much more often, he says.
Mobvoi was unwise for not insisting on a contract before working with Xiaomi on product integration. But it’s not hard to see how they made the mistake – “get it done first is our style” wrote Zhifei, on the first three months of cooperation.
“We didn’t think about it too much at the time,” a PR spokesperson from Mobvoi tells Tech in Asia, explaining why the startup didn’t insist on a contract earlier. “We were full of enthusiasm to get working.”
Founded in 2012, Mobvoi has made speech recognition one of its primary areas of research. Its main product line is the Ticwatch, a smartwatch that responds to voice commands. It has raised “millions of dollars” of funding from investors like Sequoia Capital, ZhenFund, and Google. This was the first time that Mobvoi partnered with Xiaomi.
(Correction: this post has been updated to change attribution from Xiaomi to its Mi TV team)