How to Build Creative Conversations in the Age of Voice Technology
Alexa, Siri, Cortana, Bixby and, err, Google Assistant (come on Google, would it hurt you to give it a name?) are coming to smash your screens. The rise of the very vocal AI assistants and their march into our homes via smart speakers like Amazon Echo and a host of imitators like Microsoft’s Invoke, Google Home, and Apple HomePod has led many to speculate that the future could see people – and brands – navigating a less screen-addicted existence. Add to that our growing comfort with chatting to machines, and an expectation that we ought to be able to find the information or products we want via conversation.
If this really does come to pass, it means that creatives will need to get their heads around a whole new way of working as it has huge implications for their clients. LBB’s Laura Swinton caught up with Marty Wetherall, director of innovation at Fallon Minneapolis to get his take on conversation design and getting creative with voice, audio and AI.
LBB> If the future really is to be ‘screenless’, as fans of voice-driven AI predict, what does that mean for the importance brands place on how they sound – sonic branding, voice etc?
MW> The sound of someone’s voice has a big influence on how we perceive them, so as brands begin to craft their audible brand voice to talk with consumers in this emerging context (including any audio mnemonics mixed in for flavor like GIFs in a text chat), they’ll need to consider a lasting relationship voice versus one that may be fun as a commercial VO or end tag but would get tiresome talking with it over time.
LBB> On the advertising craft side of things, the audio people (sound designers etc) are often the last thought in the production process – are they about to become a lot more sought after?
MW> Perhaps, but the dynamic nature of conversation design may not allow for the same level of control and craft as the linear soundtracks sound designers are used to. Clean, rearrangeable vocals will be the priority, but it will be interesting to see what kinds of audible ornamentations emerge as voice reestablishes itself as a primary interface.
LBB> Many people I’ve spoken to have said that voice-driven AI has massive implications for search and how people choose products online – shopping using Alexa is a bit different from using the Amazon website, as you can’t rely on ‘shelf appeal’ (online or offline), and if you don’t specify the brand then you’re relying on an algorithm to select the most appropriate brand for you… How can brands/creatives/strategists approach this challenge?
MW> Amazon has already knocked the wind out of a lot of brands with its site algorithms, and the music industry is again first in line for disruption as we struggle to tell Alexa the exact titles of our favourite songs. I expect voice to act like a new Yellow Pages where you can pay more to move up in a general search queue, but without a screen to simultaneously show paid and organic results, new relevance tech will be needed and brands will have to stay responsive to be heard.
LBB> I guess for some brands it will be about creating their own voice ecosystem (for example IOT devices, cars, maybe banks) whereas others might be more concerned about fitting into existing ecosystems. Is it clear where different kinds of brands need to focus right now, or is it still too early to know for sure?
MW> One of the coolest things about conversational A.I. is that once you’ve built yours, you can port it into multiple contexts, including the most popular ecosystems people already use. In some cases it may be necessary for a brand to build a bespoke experience for their own unique context, but overall we should expect to talk with our favourite brands wherever we happen upon a connected speaker and mic.
LBB> In this scenario, do you think it will be more important for brands to have really distinctive over-arching tone of voice/accent/speech pattern or rather to create something that can be personalised to the user?
MW> I think that depends on the brand. If you like customising landing pages for a variety of different site visitors swapping tones in social media, then a collection of voices may make sense for you and that’s certainly going to be an option. But I expect most brands will want to become distinctly recognisable on the new audio interface landscape with a signature tone and voice.
LBB> We’ve already seen agencies making chatbots to various creative ends, so they're getting their heads around AI/machine learning in the written sphere, but what are the extra things creatives and producers need to consider when developing skills/apps/brand personalities for the audio/voice space?
MW> Conversation is conversation. At Fallon, we built a parody chatbot for Arby’s Pizza Sliders that we could have spun up as an Alexa skill with only a few modifications, the big one being a decision to run with Alexa’s voice or record our own. The element of sound naturally creates the need for more creative and production decisions, and it will be interesting to see how many brands pursue recording and building responsive vocal libraries with the help of services like Lyrebird versus leaning on stock voices like Alexa to carry their tone and personality.
LBB> What sort of interesting things can you do with interactive soundscapes in this voice/AI world?
MW> We’ve only just begun to explore the potential for responsive audio experiences in this new space. As smart speakers move into our ears via location-aware headphones or earbuds, the binaural sound being developed for VR will provide 360 degrees of canvas for audio exploration. I’m excited to hear what happens.
LBB> What advice do you have for creatives and strategists who are not used to thinking 'audio AI’ first'?
MW> Consider a conversation without the benefit of visual cues. Combine the roles of copywriting and UX design to guide people through a conversational experience, anticipating their needs along the way and accounting for the way we speak versus type. Then take a step back and absorb the potential of every brand now being able to carry on one-to-one conversations at scale wherever they are and imagine the possibilities.
Genre: Creative technology