Everyone’s Talking About Voice
Thanks to everyone who joined us at The Federation on Thursday, especially the five very different but equally fantastic speakers:
- Albert Creixell, Head of Partnerships for Alexa Skills Europe
- Marcus Duffy, Head of Design at Apadmi
- Professor Trevor Cox, with the Acoustic Engineering Department at Salford University
- Emma-Victoria Houlton, Alexa Flash Briefing Producer and Founder of The Art of Podcast
- Mukul Devichand, Executive Editor of Voice for the BBC
It’s impossible to recap everything, but here are some our favourite bits:
“Customers are rapidly shifting the way they interact with technology. Why? Because Voice feels much more natural.”
Albert and his team specialise in helping brands and their developers to take advantage of this shift. Their vision? “That Alexa will be everywhere, the fabric that connects customers with technology”.
- Alexa will be an extension of your brain – On the left, analytical side, Alexa can increase efficiencies like changing the TV channel without having to look for the remote. On the right, creative side, Alexa will fulfil emotional needs like playing music to suit your mood
- Complimentary screens can make all the difference – Alexa comes in over 50 form factors and screens still have a role. For example, customers can complete most of their shopping journey by voice, but a screen means they have the peace of mind associated with visually checking through the order before hitting ‘pay now’.
- Alexa is venturing out of the home – Starting in mid-2018, Alexa will be available in BMW and MINI vehicles. This means more than music or navigation. People can use the time spent in the car to get information, shop or perform other tasks. Simple, natural tech interaction.
- Amazon’s key things to think about before you build a skill
– Value: what unique experience can you create for users on voice platforms?
– Simplicity: how will your skill simplify an action? The general rule of thumb is when an action takes more than 3 steps to complete, there’s a case for voice
– Friction: how can you solve customer problems? Account-linking, seamless payment systems?
– Frequency: is yours a skill that can slot into the every-day lives of your customers?
“Choose something the user was doing anyway, and make it easier for them.”
Already, there are countless skills or voice apps with one star reviews. Marcus Duffy, Head Of Design at Apadmi tackled this, by highlighting the lack of design so far, and what you can do about it.
Voice design is UX Design – Whether it’s a mobile app or a skill for Alexa, the process behind the design stage is very similar. So instead of wireframes, create sample dialogs. And whilst a user journey is just as critical, use a detailed voice script rather than visuals to brief the developers.
- User testing is still critical but needn’t be costly – Use simple table reads or ‘Wizard Of Oz’ to test your utterances and responses, as early as you can in the process.
- A lot of voice design is copywriting – You’ll need access to top writing skills for voice design. A few tips:
– Write for the ear rather than the eye – use contractions like “I’m” instead of “I am”
– Add variety to your responses to bring humanity to repetitive tasks
– Ensure user interactions can be said in one breath…
– …but balance brevity with friendliness. Too short can sound cold and robotic.
- Tailor the app to the user – Create novice prompts for new users that help them learn the skill. Equally, find ways to short cut the conversation for frequent users. Take a recipe skill, for example: “I can help with breakfast, lunch or dinner. What meal are you cooking?” vs “What kind of meal?”
- Use the cool audio enhancements – Read up on Speechcons, Earcons and especially Amazon Polly to give your skill the personality that fits your brand.
“How you sound is just as important to your identity as what you wear”
Enough tech! As a professor of Acoustic Engineering, Trevor shared a quite different view of voice, exploring what our voices can reveal about us.
- Voice indicates social class – How do you say “scone”? Does it rhyme with ‘stone’ or ‘gone’? We attach such strong emotions and ideals to our voices that while this is a very weak marker of social class “our heuristic brain” assumes that ‘the other’ pronunciation is wrong – you’re either too posh or too common!
- The voice is flexible – We can change our voices at any time, and in doing so, we change what we’re portraying about ourselves to others. Margaret Thatcher deliberately dropped her pitch throughout her leadership to assert her authority
- We attach ideas to certain voices – In a provoking study of male heterosexual actors who play homosexual characters, Trevor revealed that they often heighten the pitch at which they speak closer to that of a woman. But measurements of the genuine pitches of straight and gay men show no actual difference
“70% of users are listening to more audio content now they have a smart speaker”
Emma started to produce flash-briefings when she recognised the power of short, snappy audio content on voice platforms. These bite-size news pieces slot into the daily lives of voice lovers and provide a constant touchpoint for brands. Here are a few of her suggestions on creating top flash briefing skills:
- Keep it short – Flash briefings typically only allow around 90 seconds’ worth of recording, as users prefer concise voice content
- Avoid adverts – NetElixir’s recent survey showed that only 30% of people would welcome sponsored content or advertised product suggestions through voice assistants
- Leave out the robots – Text-to-speech is no-go. Use an actual voice for your pre-recorded content and you’ll stay away from bad reviews
- Update them daily – People expect flash briefings to be very regularly updated, especially when the majority who build them are news outlets.
- Upgrade your equipment – One of the major issues with flash briefings can be the audio quality. Keep background noise to a minimum, optimise it accordingly and ensure it’s in the best format
“Voice in an opportunity – the chance to have a conversation with your audience”
Finally, we heard from the BBC about how they are using voice to inform, educate and entertain just like any other media platform, and how voice has still plenty of evolution ahead.
- Richer game design – Voice has fantastic potential in this field. There are plenty of games out there, but they need to be refined to offer the level of immersion that traditional games do
- Deeper content – To compete with major news channels, voice must offer the option to go deeper than the 90 second clips available. Listeners will want entire articles
- Apply device context – Voice assistants should evolve to apply context to the user experience depending on what device they’re speaking to, with responses changing accordingly
- Utilising machine learning – This will allow a truly unique experience for each interaction
- Improve audio quality – Established broadcast stations currently have the upper hand in terms of audio quality
And that’s it!
It was a brilliant night – thanks so much to all our speakers, attendees and everyone who helped to make this event. Why not see if you can spot yourself in our pictures gallery…