Will 2018 Be a Tipping Point for Voice?
In this regular Street Fight feature, local marketing gurus David Mihm and Mike Blumenthal kick around some of the biggest ideas affecting the local search ecosystem and the broader industry. Send us an email or leave a comment if you have specific topics that you’d like them to touch on in future columns!
David: Hey Mike, hope you had a great holiday break! Mine was excellent. I’m hard-pressed to remember a Thanksgiving-Christmas combination where I ate quite so well.
Mike: It’s been incredible to have 4 days off during lake effect snow squalls. I got some great cross country skiing in between the incredible family holiday celebrations. All is good! And my brisket was great and my daughters babka was to die for.
David: Well, sadly, it’s now back to the grind. I thought maybe this week we could talk about our views on voice devices — they were ubiquitous during NFL, NBA, and NCAA commercials last weekend. The battle between Home and Echo was more interesting than most of the NFL games I watched.
I have to think 2018 will be a tipping-point year for voice adoption based purely on how much money Amazon and Google have spent advertising over the holidays.
Mike: Not being a big sports fan I didn’t notice their ubiquity. But as you know I am less sanguine about intelligent voice microphones as the core metaphor for human machine interactions over the long haul. While the technology is interesting the microphones seem too much like a command line for me to think that they will dominate for the long haul.
David: Well, we’re at such an early stage of the voice interface that it’s premature to call it a “command line.” It’s definitely rudimentary at the moment (in addition to being plain ol’ rude) but with machine learning I think things will start to feel a lot more natural in just a couple of years. I mean if Facebook robots can invent their own language, surely Google, Amazon, and Apple can adapt their responses to English (and other human languages).
Mike: So by “command line” I am referring to the need to memorize actions. That means for a human to learn more than a few actions on any give device will be very difficult. It will work for the types of commands that serve Amazon well like “order more Tide” but from that point on it becomes more problematic for most humans to remember the many commands need to get things done.
David: We’re already seeing that–for a certain demographic, voice assistants are little more than personal meteorologists.
There may only be 6-7 “commands” that we use regularly with voice, though I’m willing to bet that as people get more and more accustomed to the interface, the horizon of possibilities expands.
I don’t understand, why that still doesn’t qualify a huge breakthrough in your mind. After all, most people only use 6-7 apps regularly, but I’m sure you’d characterize the iPhone interface as a huge breakthrough relative to its Blackberry and flip-phone predecessors.
Mike: The reason for me is that the metaphor of the iPhone interface clearly can be expanded to cover 5,6 or 600 apps in a natural way and allow the user to “peel the onion” as they dig deeper into the functionality of the device. I fear that with voice users will confront a “wall” where they can not move beyond the few commands that they have learned.
David: I think voice will work the same way, though. People will discover new ways to use their voice devices (or new Skills for Amazon users), and the really cool ones will spread virally, just as new apps are developed and spread on smartphones. Whether or not they’re regularly used by the majority of users is a little secondary to me.
For example, I doubt I’ll carry a single airline app on my smartphone once voice integrations are further built out:
- “Hey Siri, send my boarding pass for my flight tomorrow to Apple Wallet.”
- “Hey Alexa, see if you can get me upgraded to First Class on my flight tomorrow.”
- “OK Google, find me the cheapest nonstop flight to any major airport in Europe in February.”
I only use airline apps on my smartphone on the day before or day of travel, just as I’d only use those voice commands around the same time. That doesn’t mean I won’t be able to discover them on voice devices — the airlines will promote those Skills in the same way they promote their apps.
Mike: Those are ideal use cases but for an assistant to be truly effective it needs both natural language processing AND hooks into a ton of services.
David: Agreed, both NLP and a plethora of service hooks are essential. Before we get to NLP, though, there are two additional fronts where voice needs to mature: recognition and processing speed. Where are we at the moment on the former–95%? I’m sure that last 5% will mirror a golf handicap, where the closer you get to scratch the harder it is to shave strokes. But at least in my case, I’ve been pretty happy with voice recognition on all three platforms.
It’s the processing that needs to get faster (or at least be asynchronous, the way Google Suggest works). Too many times I’m stuck waiting multiple seconds for a response, which is just as frustrating as typing the same query or command into my phone. In that regard, voice is not yet a quantum leap forward for user experience.
Mike: While speed will help with recognition and the appearance of intelligence, I would point out that speech recognition is not natural language processing which is what we need to make voice really work.
Essentially Alexa and Google Home are translating commands and then slotting them into the limited framework of actions that are supported. I think the technology to get to where you want to go is a ways off still and not a 2018 event.
David: Here’s where we disagree, I guess. I think we’re very close to tipping-point support for the structured “commands” we perform currently in frequently-used smartphone apps. The actions we perform in these apps are repetitive:
- “Siri, order me a Lyft in 10 minutes.”
- “OK Google, how old is Tony Parker?”
- “Alexa, order me a Yellow Curry for delivery from Tarad Thai.”
Mike: Yes we do disagree. It might be time for another beer bet as to how soon speech and voice assistants take on the utility you describe to broadly used.
David: Great, a chance to earn my money back from our Posts bet, which is not looking good for me!
Mike: On another more pragmatic front for marketers, fragmentation is a big issue as well. For us as marketers to be able to leverage a new technology in local there needs to be enough penetration amongst users AND enough market share presence to achieve broad gains for our clients. Due to the fact that Apple controls so much of the speech market in the US, It could be years before a platform that is useful for local businesses matures even if the technology is broadly adopted.
David: While voice does fragment the market to some extent, one of Google’s best strategic moves ever has been its preparation for the world of voice with its desktop OneBox / Knowledge Graph / Featured Snippet answers. It’s pretty clear that results are going to be identical, just presented in a different medium.
So the real fragmentation is going to be the additional installed userbases that Apple and Amazon provide, in my view.
Mike: Google certainly has a market leading knowledge graph that will give them an advantage in all of this.
But my list of concerns go beyond the need for natural language, fragmentation and industry leading architecture. There are significant security AND privacy issues. My wife said to me: “No Google Home in our home!”. I think those factors will be issues for more and more people going forward. It will definitely slow adoption of Google and Alexa down and give Apple time to make a real play in the market. Probably with Airpods though and not an intelligent microphone.
David: I’m unconvinced that privacy matters much to my generation or millennials, although it’s certainly stopped me from being an eager adopter. I’d prefer a remote control that turns on the listening feature of my voice device.
Failing that, perhaps artisan cone-of-silence Echo accessories are going to be Portland’s top-selling gift for Christmas 2018.
Mike: I hope you are wrong on the privacy front.
But perhaps we agree more than we disagree and it’s really just the time horizon where we have the squabble. As Bill Gates said, “[w]e always overestimate the change that will occur in the next two years and underestimate the change that will occur in the next ten. Don’t let yourself be lulled into inaction.” That is most certainly true in this case.
After more than a decade in local search, David Mihm now runs Tidings, an email newsletter platform for small businesses that leverages their everyday social media activity, and his own weekly newsletter, Minutive. In 2012, he sold his former company GetListed.org to Moz, helping over 3 million businesses get better visibility in Google and other search engines. Along with Mike, he’s a co-founder of Local University.