When McDonald’s spent over $300 million on big-data-crunching startup Dynamic Yield earlier this year, the move came as something of a surprise. The follow-up should not. Today the Golden Arches announced the acquisition of Apprente, a voice AI system focused on fast-food ordering. It’s a niche, but it just paid off.
Specific terms of the deal have not been disclosed. But the synergies are at least more immediately understandable. Apprente’s speech-based artificial intelligence deals within the relatively narrow confines of quick-service restaurants. As with Dynamic Yield’s decision engine, which switches up menu items based on what it thinks consumers want at any given time and location, Apprente’s ultimate goal is to increase the speed of any given transaction. Anyone who’s had to repeat their order into a squawking speaker knows that pain.
Apprente calls its technology “sound-to-meaning,” in contrast to “speech-to-text.” The distinction, other than having a nice ring to it, is that, unlike many voice AI models, Apprente says it does not transcribe what the customer says and then infer meaning from that transcript. It goes directly from speech signals to result. “The company believes this provides a better approach for customer-experience-related use cases, particularly in noisy environments such as restaurants and public areas or in cases where customers tend to use colloquial, poorly structured language, resulting in low-accuracy speech recognition,” Raúl Castañón-Martínez, a senior analyst at 451 Research, wrote in a recent report on voice assistants. In other words, in environments like a drive-through, where the stakes for fast-food restaurants are impossibly high: The majority of the 68 million customers McDonald’s serves every day never leave their cars.
Because it plays within the confines of a menu with a limited number of items and a fairly predictable interaction paradigm, the Apprente system may well have an easier time making those interactions sound more natural than more free-ranging voice assistants do. Drive-through-window exchanges often involve accents, a challenge for even the most sophisticated voice AI. Order customizations can require hunting and pecking by the operator to input, potentially slowing down the line. And Apprente has touted its ability to incorporate order changes on the fly more quickly than its human analog as an example.
But does it really work? As with Dynamic Yield, the proof may be in the purchase. McDonald’s has historically not been trigger-happy with its acquisitions. It was apparently pleased enough with how Apprente’s voice assistant performed in real-world trials to snap it up. And unlike Dynamic Yield, which has remained a separate corporate entity wholly owned by McDonald’s, Apprente will come entirely in-house. That makes this as much a defensive acquisition as anything; Burger King and Wendy’s are now going to have to cook up, or acquire, their own voice AI. Not only that but the Apprente team will become founding members of the newly formed McD Tech Labs, a Silicon Valley–based group that will work on better drive-throughs—or faster ones, at least—through technology.
“Building our technology infrastructure and digital capabilities are fundamental to our Velocity Growth Plan and enable us to meet rising expectations from our customers, while making it simpler and even more enjoyable for crew members to serve guests,” McDonald’s CEO Steve Easterbrook said in a statement, referring to the strategic vision he laid out in 2017.
On that last point, an Apprente press release last fall touted that its AI “offers a more consistent and pleasurable customer service experience with its virtual agents never sounding tired, annoyed, unhappy, or angry.” McDonald’s declined to comment on whether its efficiencies would also result in fewer human jobs.
A McDonald’s spokesperson also described it as “premature” to discuss plans for deployment and did not share details about how extensive the testing has been. In other words, it’s not clear when the voice on the other end of the McDonald’s order kiosk will be a robot or if you’ll notice once it is. (Hopefully McDonald’s will tell you up front, given the subtle creepiness of talking to a machine unawares.) If Dynamic Yield is any kind of point of reference, McDonald’s made that acquisition in March; today it’s in over 8,000 of the company’s 14,428 US restaurants. It’ll be in “nearly all” by the end of 2019, according to a McDonald’s press release.
social experiment by Livio Acerbo #greengroundit #wired https://www.wired.com/story/mcdonalds-acquires-apprente-voice-ai