with its Google Home smart speaker, as did Apple with its Apple HomePod. The big tech companies were focusing their own R&D resources on the refinement of speech recognition technology using artificial intelligence software trained on big data sets.
For Nuance, the consumer device market was a closed ecosystem that was tough to penetrate, says Daniel Ives, a senior technology analyst and managing director of equity research at the financial services and investment firm Wedbush Securities.
“They were kind of on the outside looking in,” Ives tells Xconomy. “They had the right strategy, but so did [Amazon founder Jeff] Bezos and [Apple CEO Tim] Cook.’’ (Wedbush makes a market in Nuance shares, executing buy and sell orders from clients.)
By 2017, Nuance reported in SEC filings that its revenue from consumer device makers was continuing to decline “as a result of the consolidation of the mobile devices industry.’’
Last year, Nuance announced plans for a major restructuring, including the spinoff of Cerence, which became the sole remaining component of a larger Nuance division once called Mobile.
Going forward, Nuance will focus on its two other top-earning units. The healthcare division offers Dragon Medical, which helps clinicians use speech commands as they document office visits and make entries in electronic health records. The company’s enterprise division serves companies that want to build a conversational interface with consumers, such as customer service chat features.
“Our wheelhouse is B2B,’’ Mack says.
Wedbush analyst Ives, for one, approves of the Nuance plan.
“This is an exciting new chapter for Nuance after many speed bumps,’’ Ives says.
Cerence’s turn to compete with big tech
Arguably, conditions are very different for Cerence in the auto industry than they were for Nuance in mobile devices and IoT. For one thing, the “devices’’ that operate using Cerence’s technology are not owned by big tech companies, but by automakers who have their own business agendas.
Cerence inherits Nuance’s relationships with all major car manufacturers or their top suppliers, going back decades, the company says. Its white-label software has been built into a total of about 280 million cars made by all major auto manufacturers, including BMW, Ford, GM, Toyota, and Volkwagen. That total includes the more than 45 million equipped last year.
Mack says Nuance’s work with automakers began in the late 1990s and early 2000s. Ford was one of the company’s original flagship customers, and its launch of the Ford SYNC system in 2007 was an “industry watershed moment that put voice on the map,’’ Mack recalls.
“We were behind that and we’ve ridden the wave ever since for automotive purposes,’’ he says.
Cerence has spread roughly 450 professionals out to the car-making capitals of the world to help manufacturers create custom conversational AI features that take into account factors such as the acoustics of their vehicles and the functions their customers expect. The company says it’s prepared to serve automakers selling to international markets. Its speech recognition technology covers 70 languages and dialects, from English and Spanish to Mandarin and Shanghainese. Of its 1,300 full-time employees, about 700 are engineers.
Unlike companies that make their trading debuts after an IPO, Cerence won’t begin with a fresh infusion of cash from investors. Instead it will take on loan debt of about $425 million, and pay roughly $314.2 million of that money to Nuance. Mack says Cerence is absorbing a share of the debt Nuance was carrying when the Nuance automotive unit was part of its business. Cerence will retain about $110.8 million from the loan for its own operations. Nuance will retain no ownership stake in Cerence.
Dhawan says Cerence will be able to pay down part of that $425 million debt every year, because it has positive cash flow. Revenue of $277 million and net income of $5.9 million was attributed to the Nuance automotive division for fiscal year 2018, which ended Sept. 30. Cerence is estimating that its 2019 revenue will reach $308 million to $310 million.
“Our balance sheet is very strong,’’ Dhawan says.
That said, Cerence acknowledges in its SEC filings that competing with “large technology companies that have significantly greater financial, technical and marketing resources than we do’’ could affect its prospects. For example, such companies might try to gain new business by offering their products at “low or unsustainable cost.’’ They also compete for engineers skilled in the development of AI software. Dhawan says Cerence is doing well in retention and hiring by offering engineers the chance to make significant contributions to the creation of new products.
Another important factor for Cerence is that it must design its products to work alongside the software and hardware of big companies such as Amazon, Google, and Apple.
Automakers, as well as Cerence, understand they must “bring into the car the digital life of the consumer,’’ as it already exists at home and at work, Dhawan says. That means, at least to some extent, providing access to the virtual assistants that consumers are used to—such as Amazon’s Alexa, Apple’s Siri, and Google’s Assistant.
“They don’t want the car to be a separate island,’’ Dhawan says.
Drivers already plug their phones into their connected cars and use their existing apps for functions like navigation and music. Amazon has created its own small device, the Echo Auto, that drivers can park on their dashboards. Beyond that, Amazon,
