Nuance is the world’s largest voice recognition technology company. Over the past ten years, Nuance has experienced countless ups and downs: Apple, Google, and Samsung have been in the hands of most voice companies. Afterwards, he was repeatedly faced with market crash, technical bottlenecks, customer loss, turnover, missed the best selling time, and so on. Eventually he had to transform from a technology provider who only focused on the vertical deep plowing algorithm to a company that expanded horizontally across various industry solutions.
 "Nuance Dependence" and "Nuance phobia"Lei Fengnet (search "Lei Feng Net" public concern) has repeatedly reported on Nuance's in-depth articles, pointing out that Nuance was first known to everyone due to their cooperation with Apple. After Siri, who was known as the next-generation interactive system, adopted their voice technology, the company, whose name was not seen and voiced, was instantly exposed by major media. In fact, before cooperating with Apple, they have already maintained cooperative relations with a number of well-known manufacturers. Their technology has been applied to mobile phones, televisions, and automobiles, among which Samsung's voice assistant, S-Voice, also uses Nuance's technology.
As the entire voice technology market Nuance family is the only one, the technology far exceeds its competitors, making the customer base highly dependent on them.
Nuance's CEO, Paul Ricci, is a famous horn in Silicon Valley. During his tenure he led 60 mergers and acquisitions, Ricci was extremely adept at using litigation to weaken those innovative competitors in order to acquire them at very low prices or to bankrupt them. Dave Grannan, CEO of startup Vlingo, said:
Competing with Nuance is like an infectious disease. We are leading the way in direct competition with them. But when you think life is full of beauty, hey! This is the end of a good day.
McCue, the former CEO of another start-up Tellme, once mentioned that many companies can't withstand the pressure of Nuance and finally give in, and that such things are still happening.
At that time, Nuance had the absolute status and right to speak in the industry, and any companies associated with it had to surrender to it, and the giants courteously courted them.
However, this arrogant voice giant is facing an unprecedented crisis, but he quietly do not know.
 Speech empireAfterwards, Nuance was faced with a series of fatal problems in a short period of time, and the problems were concatenation, a rupture, and the whole link could collapse.
Technical ceiling
Nuance's speech technology is based on statistical inference methods, focusing on phonemes (voices of syllables) and context to identify words. In terms of recognition, Nuance's algorithms and patents lead other competitors. However, speech recognition technology is not unattainable. Especially after 2010, patents and algorithms are playing a less and less important role in the context of speech applications. The pure recognition rate of each voice gradually converges to the same level, and the gap is no longer sensible. If we continue to deepen the recognition technology, the road will be narrower and narrower, and the cost will not be directly proportional to the effect of improvement.
Loss of customers
At this time, deep learning was unprecedentedly fierce. It was widely used in speech technology, eliminating the need for many speech model creation links, making the difficulty, cost, and time all drastically lower. A large number of manufacturers poured into this industry and Nuance was in the same dimension. Start the competition. At this time, companies such as Apple, Samsung, and Google have begun to focus on developing their own voice technology in order to get rid of their dependence on Nuance. Nuance's technological advantage is gradually absent and customers are slowly losing.
Core talent is frequently dug
Losing customers began to gather talents and develop their own voice products, and digging Nuance corner is the most ready-made method. As a strategic partner of Nuance, Apple has the most high profile of its man-occupation. Apple quietly established a voice technology research and development team at Nuance's hometown of Boston. It not only absorbed voice technicians who serve Siri, but also put Nuance’s senior executives and Senior speech researchers were recruited, including Larry Gillick, formerly vice president of research and development at Nuance, and Gunnar Evermann and Don McAllaster, senior voice scientists.
At the same time, in order to avoid Nuance's patent obstacles, Google poached co-founder Mike Cohen of Nuance for 10 years to work as a “speech leader†to develop corresponding speech recognition technology. Interestingly, Hugo Barra, Xiaomi’s global vice president, was also the product manager of Nuance and was recruited by Google in 2008.
The loss of core personnel means that Nuance's technical barriers have been disguised as disintegrated, and the reliance of large companies on it has been decreasing. This is equivalent to Nuance, who relies on technical output to obtain revenue.
Missing the best sales period
Jobs had made a request to acquire Nuance, but was rejected by the Nuance CEO. The industry believes that they apparently missed the best selling period.
Later, when Nuance was in a dilemma, Samsung and Baidu also showed their intention to acquire Nuance. Voice interaction instead of multi-touch has become the consensus of people in the science and technology circles. Voice technology is often more important than its actual value for enterprises with over 100 million users. However, the final acquisition has not progressed smoothly. Analysts pointed out that companies such as Samsung mainly consider these two issues: One is the price/performance ratio, and Nuance employs more than 15,000 people. If the acquisition transaction scale will be more than US$ 50,000 to US$ 6 billion, potential It is a question of whether or not the acquirer's large Nuance merger is worthwhile. Secondly, Samsung initially conceived that by acquiring Nuance, it would restrict the development of Apple and Google in related businesses to a certain extent. However, the reality is that each company’s own voice technology has matured, so “when Nuance is taken, you can control it with Nuance. If the phrase "cooperative enterprise" was established before, it has now become a false proposition. Potential acquirers have to give up.
Looking back at Nuance’s experiences over the years, his decline is inseparable from the word “technologyâ€. First of all, Nuance himself encountered the ceiling of voice technology and it is difficult to make major breakthroughs. Secondly, with the deep learning tide, the scrambler who used the new algorithm far exceeded Nuance's expectation and became overwhelming. At this time, the time left for Nuance seems to be running out.