With the likes of Alexa, Google Home and Siri, voice assistants are more popular than ever. Capable of making our lives easier in a flash, the devices featuring these AI are fast becoming an indispensable part of our everyday. But what does the future hold for this tech? Dave Waller explores the power of voice.

Way back in the 1960s, MIT professor Joseph Weizenbaum introduced the world to Eliza, a responsive computer system. Eliza was hooked up to an electric typewriter, on which people could fire off all manner of personal concerns. The computer would respond sensitively, to elicit more detail from its users. It felt like a real conversation. Eliza was essentially harnessing a grammatical trick – rephrasing what you’d said and throwing it back at you as a question that appeared empathetic – but that didn’t matter: Weizenbaum saw users becoming emotionally attached to the machine. And so he pulled the plug. “When the computer says, ‘I understand,’ it’s just a lie because there’s no ‘I’ and there’s nothing that understands anything,” he said. “That’s just wrong.” 

But the machines weren’t left unplugged for long. With powerful AI at the helm, the modern descendants of Eliza no longer rely on sleight of hand, nor the written word, to win our trust. In the first quarter of 2018, Google sold 3.1 million Google Home smart speakers, Amazon 2.5 million of its Alexa-powered Echos, according to analyst Canalys. And they’re not just helping us with our emotional issues: they’re increasingly capable of ordering pizzas, running home entertainment systems, and organizing a calendar. Amazon has now released an Alexa-powered microwave. And in September 2017, the 21st season premiere of South Park activated Alexa and Google Home devices around the world to set alarms and create shopping lists that included items like ‘hairy balls’. It’s not clear what Weizenbaum would have made of that. But if he saw the furore it caused on Twitter, he’d surely note the emotional power. And probably reach for the plug again. 

And these machines are still only learning to talk. 

The new mobile

Lee Mallon, founder of Bournemouth-based digital agency Rarely Impossible, remembers the moment he first realized the power of voice. He’d imported an Alexa system from the States four months before its UK launch, but found himself using it only occasionally, for setting kitchen timers and playing the odd song. Then his kids started using it. Every day. 

“As adults we get frustrated because we expect something to work as well as mobile,” he says in an early morning call, his children clattering around in the background. “We’d try two goes at an instruction, and then go ‘oh for fuck’s sake’ and just use Spotify on the phone instead. But for my kids this was their first bit of tech. They’d ask the same thing 10 times and just tweak it until it was right. They played the ‘I Like to Move It’ song from Madagascar so much that Alexa even began to understand my young son mumbling. That’s when I said: ‘Holy shit. This is their mobile.’”

When Amazon launched its smart speakers in the UK, Mallon gave a few to friends. The pattern repeated. A friend’s daughter once got in the car and said: “Car, take me home”. But while these voice systems are still most popular among children and the over-60s (only an estimated 3% of people have attempted to buy things via voice, according to Paul Armstrong, head of London agency Here/Forth), this may well be about to change. The tech is becoming more accurate and more useful, and it’s creeping into daily life. While predictions around future tech habits tend to be sketchy – “Anyone who says how it’ll play in five years is talking crap,” says Mallon – voice does seem a no-brainer. Speech is natural and fast. Smart speakers are cheap and immediately easy to use. And the companies pushing it have deep pockets – and are taking it very, very seriously. 

“The beauty is people don’t need to leave the platform they’re on in order to complete the action – unlike other popular brand platforms, like Instagram, they can just see an ad for a red ASOS dress and, without as much as a swipe, ask Alexa for it.”

“Amazon is pumping billions into this every year, so it’s not messing around,” says Mallon. “It wants to sell you shit. It has every product in the universe. If you need bin bags, with voice you’ll never need to go out and get them. You won’t even have to search for them. The other company driving it is Google. Our minds are being blown by the things we’re privy to, but that’s just the stuff these companies are sharing now. What about the other things still in R&D? Do brands need to think about this?” 

That’s obviously a rhetorical question. It’s now 10 years since Dominos pizza launched its Dom chatbot platform, an unlikely pioneer of voice on the web that allowed people to speak their pizza orders. Plenty of other major brands have their own experiments well under way. In May, Lego launched its Lego Duplo Stories, in which Alexa guides the user through an interactive story about animals and vehicles, suggesting Lego bricks and props to use to play it out. In June, H&M’s homeware brand launched H&M Home Stylist, a Google-powered voice assistant that provides styling suggestions and mood boards, with an exclusive human voice. 

With the barriers so low, it’s just going to spread. Amazon already has your bank details, so people will soon be making charity donations from their sofa, just asking Alexa to donate £10 having seen an ad for, say, Children in Need. The beauty is people don’t need to leave the platform they’re on in order to complete the action – unlike other popular brand platforms, like Instagram, they can just see an ad for a red ASOS dress and, without as much as a swipe, ask Alexa for it for Christmas. Done.

Really? A kebab?

But if brands are already experimenting, they’re also starting to make some familiar mistakes. Last year, when Ocado launched its Amazon Alexa provision. Enabling people to add to their shopping lists by voice, it proudly announced an ‘open-collar’ approach – resulting in responses like “okey dokey”, and “over and out”. While cutesy phrases like this may seem a good idea in a pitch meeting, they can quickly become irritating. 

“Whenever I see something that seems reasonably well controlled and leading edge, someone goes ‘fuck it’ and smashes it with a hammer,” says Julian Barra, a strategic creative who advises the likes of Virgin on undertaking its brand campaigns. “If the past is illustrative, voice will be another one of those: ‘I’m going to call my virtual assistant a c*nt and I want it to be able to deal with it.’ It may be funny for a bit, but the industry will be convinced they should all gallop in that way – suddenly doing radio ads that make Alexa order a pizza you didn’t want, or trying to act like your best mate. Brands will try too hard and it won’t work. Imagine:

“‘Google, is that kebab shop still open?’ 

‘Really Julian? A kebab?’ 

‘You can fuck off actually, Google. I want a kebab.’”

“Voice should be no different to any other branding. It’s a technology that becomes part of a bigger brand voice and it needs to be managed – not just ran at with your hair on fire, in your big trousers and clown shoes.”

Julian Barra, creative strategist

Barra is adamant that voice is going to become increasingly important to his clients, just like choosing the right person to narrate a radio ad, or writing the right script for its call centers. And beyond thinking hard about the user experience, and being careful not to invade that user’s protective circle of trust, he says the key lies in being consistent. He should know what he’s talking about, having created the iconic nodding dog campaign for Churchill Insurance. 

“That was a group of people who were very, very clear on these matters, creating one voice that ran from ads to social media to call centres,” he says. “Straight up, no nonsense, get shit sorted for you, all seamless. Voice should be no different to any other branding. It’s a technology that becomes part of a bigger brand voice and it needs to be managed – not just ran at with your hair on fire, in your big trousers and clown shoes.”

Within five years, says Mallon, every brand will have its own audible logo. “Voice isn’t just how Alexa communicates,” he says. “Think audible experiences. Virgin may want a Richard Branson-esque voice because he carries brand value. And when they have a podcast, is that Branson voice the host? Create that tool. If you use the same voice in your social ads and TV spots, it fires the neurones that make the connections.”

Building your skills

Not that voice is going to be easy for brands to get their heads around. The danger is that, for now at least, the real power of these platforms rests with the Googles and Amazons. The first step for most brands will be to wrap their heads around voice search, ensuring its results come up when a user searches relevant services in the area. Amazon uses Bing for its search data, for example, so brands will need to ensure they’re visible there to show up on Alexa. Similarly, when someone asks Google Home for the best restaurant nearby, it takes the top three from its search listings. “If I was a brand or business in a local setting, my next six months’ investment would be incentivising people to leave Google reviews,” says Mallon. 

But the gatekeepers aren’t just there to help you get noticed. In 2017 Amazon famously took all the data it had gleaned from selling other leading battery brands, and launched its own Basics range. Sales among those other battery brands were suddenly destroyed. 

Hence the big push in 2019, as well as creating voice search, will be for a brand to build its own Skills, brand-unique sub-sections of Alexa that allow users to specify its products, thus cutting Amazon out of the process. “The companies that understand their offering is a platform, not a product or store, will build a Skill,” says Mallon. “Deliveroo customers won’t ask Google where to get food. They’ll use Alexa to ask Deliveroo what the best chicken is. For many companies it’s logistically painful to roll out something that’s new, but if you’re a Deliveroo, Uber or ASOS, you can move quickly and your user footprint allows you to dominate that space.”

One of the first tasks for brands will be to raise awareness of what the hell is going on here. Increasingly we’ll see brands inserting an Alexa or Google instruction in their TV, YouTube and social ads, instead of nudges towards Instagram or Facebook, educating users how it works. “Conversion stats will be abysmal,” says Mallon. “The ROI will be low as people find their feet with it. But in 2020 things will be different.” 

“If I was a brand or business in a local setting, my next six months’ investment would be incentivising people to leave Google reviews.”

Lee Mallon, Rarely Impossible

Closing the gate on the gatekeepers

Further down the line, the task will be for brands to try to cut out the likes of Amazon or Google altogether. They may be able to employ user data to know when they’re about to run out of batteries and proactively approach them to see if they need more, so the user never needs to ask Alexa. Or, for some brands, moving beyond Skills within Alexa to create their own independent platforms. Dublin’s Voysis and California’s SoundHound are among the startups designed to offer established retailers and clothing brands their own provision. “It’s just like people starting off writing on Medium or Twitter, but eventually saying they need to host their own site, to control it,” says Mallon. “Having the most liked page on Facebook is amazing – until they change the algorithm.”  

Such independent provision will help fuel the other big trend, the spread of voice beyond the smart speaker. But as the technology advances, and the bombardment from ill-considered ‘hairy ball’ command campaigns spreads from the world’s living rooms to our cars, phones and smartwatches, the language of the brand isn’t the only question being raised. Other more nuanced concerns are emerging too – like whether kids should thank Alexa for its help. 

“Alexa is command-driven, so people are worried kids who are using it are now being rude to adults. So should kids thank Alexa? My view is no, because it’s a computer, not a person. But now if you say thank you, Alexa will say: ‘That’s very nice of you.’ The powers that be in tech want to make it as emotional as possible, because it’s how you build trust. There’s a reason that Alexa’s called Alexa.”

Our call ends with chaos unfolding in the background. His daughter has momentarily locked herself in the cupboard. Moments later, a text comes through. It’s from Mallon.  

“Apologies, unfortunately Alexa hasn’t created a kids be good Skill.”

Not yet…

Ready to rock?

Fancy a chat about this? Or maybe we can help with our advisory services or get in touch!


Join hundreds of your peers and get your weekly fix of marketing street knowledge from the Rockstar CMO community.

Share this article