Mobile Apps Need Better Speech Interfaces
Voice assistants are everywhere. I have one in my car. (Its speech recognition is far less than perfect, but what should I expect from an 11-year-old model? No doubt, newer Honda Accord models would have some of the better engines available today.) I have another voice assistant on the smart TV in my living room. I have yet another one (Alexa) on the Amazon Echo Dot in my bedroom. And I have Siri on the Apple iPhone that I take almost everywhere I go.
In a funny digression, this weekend I completely befuddled my iPhone’s assistant when I was having fun with my dog and playfully called him “Silly” when he failed to bring back the ball he chased during our game of fetch. I wasn’t aware that I had activated the voice interface, and Siri sat idly by waiting for my request, running down my iPhone 7’s already-failing battery. I wasn’t aware that Siri no longer required the word “hey” before her name to invoke her.
I mention this because this scenario is one of the many considerations raised in this issue’s feature on mobile app voice development, “Best Practices for Bringing Voice Assistants to Mobile Apps.” When developing voice interfaces for mobile apps, it is important to consider the context in which the app might be used. While Apple developers probably couldn’t have anticipated that I would be calling my dog “Silly,” the company could have reformed its speech recognition engines to filter out words that might sound like “Siri” but have far different meanings.
Apple rolled out Siri in 2011, and its speech recognition has improved little in all that time. Some might even argue that it’s gotten worse, though that is probably more of a perception than a reality. But then again, when reading the text transcripts of voicemails I receive or the text messages that my friend Bob sends to me by dictating them into his iPhone, I sometimes wonder if Siri is hard of hearing.
Apple has also fallen behind many other speech app developers by not enabling Siri to integrate with many of the other applications on my phone. I’d love to be able to use Siri’s voice interface to interact with my insurance company through the GEICO mobile app, pay my bill on the Spectrum mobile app, cash in my rewards in the Staples mobile app, or initiate a trade in my Yahoo fantasy football app. But alas, Siri can’t do any of that.
Don’t get me wrong, though. I like Siri. I am even moving off my Windows-based PC and switching to an iMac because it contains a Siri voice interface while Microsoft completely killed off its Cortana assistant. Nevertheless, the folks at Apple HQ in Cupertino, Calif., might benefit from reading our feature. I suspect that you will as well. After all, now is the time to consider how your company could implement or upgrade its voice assistant mobile app interfaces to deliver exceptional user experiences that delight and engage users, the article concludes.
AI is, of course, upending the entire mobile voice technology landscape, but its real impact in the larger speech industry is perhaps most obvious in the area of speech analytics. As our cover story, “Speech Analytics Expands Beyond Voice,” points out, with digital communications replacing voice-based channels for customer interactions, speech analytics is undergoing perhaps its most radical change ever. AI, the article says, is enabling faster insights from a much larger pool of information that includes not just voice but also text, images, video, and more.
“Contact centers can no longer rely on phone conversations alone to serve customers,” Dave Hoekstra, a product evangelist at Calabrio, says in the article. “They must leverage data from multiple digital channels, such as chatbots, voicebots, and more, to ensure a smooth customer experience.”
And that, too, puts more pressure on mobile app developers to incorporate speech technologies into their apps. And stronger analytics in mobile apps will undoubtedly lead to more robust apps with better speech recognition so that Siri will not confuse conversations with my dog with calls for it to act.
Leonard Klie is the editor of Speech Technology magazine. He can be reached at lklie@infotoday.com.