Lorem ipsum dolor sit amet, consectetur adicing elit ut ullamcorper. leo, eget euismod orci. Cum sociis natoque penati bus et magnis dis.Proin gravida nibh vel velit auctor aliquet. Leo, eget euismod orci. Cum sociis natoque penati bus et magnis dis.Proin gravida nibh vel velit auctor aliquet.

  /  Project   /  Blog: The best tips for Voice design and strategy

Blog: The best tips for Voice design and strategy

from the Voice-Connected Home conference in Cologne

Photo by John Tekeridis from Pexels

First things first: what’s this post about and why should you get into it?

  • 7 & 8 May 2019: Voice innovators, designers, strategist and creators from all around Europe got together at the Voice-Connected Home conference in Cologne, Germany.
  • In this post I share my top 5 learnings from the voice experts from the teams of Google Assistant, Microsoft Cortana, Amazon Alexa, the Voice Tech Podcast and the BBC. You’ll get some advice for your voice strategy, tips for voice design, hear about the latest developments and get a taste of the future of voice.
  • Not in the mood or no time to read the whole post? I share my voice learnings and insights in smaller bits on Instagram too, every working day (smartvoicestories) and WhatsApp (send VOICE UPDATE to +31620879652).

Still here? Great!

#5 | Not actively creating a personality, is still creating a personality

This is Benjamin Dorvel speaking. And he’s responsible for the personality of the Google Assistant, so he would know. This is what he said:

“You don’t have the option to not give your voice assistant a personality. When something talks, people will always attribute characteristics to it.

This is called antropomorphism; the attribution of human traits, emotions and intentions to non-human entities. Like seeing a face in two dots and a line. We can’t help but doing this, so if you’re creating a voice assistant, you’d better take this into account. If you don’t think it through, people might come up with traits that you didn’t mean at all for your brand.

At Google, they thought this through at a very detailed level. Dorvel explained that the Google Assistant reacts differently in different countries, so her conversational style matches the culture. She has a different sense of humor and is more or less direct. At the same time her core characteristics stay the same: she’s supposed to always help you get organized, be accessible and be useful. The way she does it, depends on the language in which she’s asked.

#4 | Talking is so much more complicated than it seems

Having read my fair share of Jefferson transcripts during my bachelor’s, I can only agree with this statement. In Cologne, Alejandro Campoy from Cortana pointed out that up to now, the interfaces we use are actually very unnatural. We always have to ‘learn’ how to use them first (computers, phones, a new app).

Voice interfaces are the opposite — they are innate to us. This means a shift from designing screens, to focusing on language and communications. But because we are so used to talking, the expectations of voice assistants are huge. We need to be realistic about where we stand with voice AI at this point — the assistants are just mimicking humans. but once we get there, there will be a shift on a fundamental level.

If you’re interested in this topic, I highly recommend the book Talk by Elizabeth Stokoe. She explains all the major concepts with scripts from Friends — I wish this had been around when I was in college.

#3 | Talk with your users, not at them

You’d think this is a no-brainer, but make no mistake: talking to users instead of with them, is the default mode for so many skills and actions right now. Alexa evangelist Benoit Nachawati provided some tips on building a great voice strategy. Here’s three key pointers:

  1. Be adaptable. Let users speak in their own words. An assistant that only gets it when you phrase a question in the exact right way, doesn’t cut it.
  2. Be relatable. Don’t just let the assistant ramble on from some already-available content from your website. Think about it like an actual conversation.
  3. Be Personal. Individualize your entire interaction — voice interfaces are about 1:1 contact. Make sure it feels that way.

#2 | Voice Tech Podcast

With still so many thing to figure out in the present, it’s nice to just dream about the future sometimes. Carl Robinson, host of the Voice Tech Podcast, gave a taste of what’s ahead. These three predictions got me most excited:

  • There will be devices that can read lips or use eye detection. That means you can just look at your smart speaker, and it will know that you’re addressing it. No more wake words. And if it still feels a bit weird to talk to machines with people near, you can just mouth what you want and it will still understand.
  • Voice assistants will be with you 24/7. Movie Her-style, you’ll always have a ‘hearable’ in your ear, delivering a continuous experience.
  • The development of new kinds of content. Think about nano-casting, like Twitter for voice? I see tons of possibilities for ‘voice original’ content.

Carl will share the audio of his talk in an episode of his podcast.

#1 | Make it extremely clear what the user needs to do

When designing for voice, we sometimes forget our users are a little helpless. No screens, no help button, no overview of what’s available within the skill or where you’re at. You’re users will be clueless, so you’ll have to take them by the hand.

Nicky Birch, lead producer for voice at BBC Research & Development, shared some in-depth tips to help you with this.

  • Be direct. Don’t ask: “Tell me when you’re ready”. Ask: “Are you ready?”.
  • The beginning of the skill is critical. On the one hand it should not be as broad as “What can I do for you?”, because your user doesn’t know yet what the assistant can do. On the other hand, it shouldn’t be a manual either. Try to blend things in naturally. So don’t go into instruction mode with: “Say ‘go forward’ when you want to go forward”, but just let the assistant ask “Do you want to go forward?”.
  • Add a ‘helper’ figure to your skill or action. When the user seems stuck, this helper can point him or her in the right direction.

The BBC published actually published a white paper last week with all their learnings. Read it and share it!

Liked this post? I post learnings and insights on voice like this via Instagram (instagram.com/smartvoicestories) and WhatsApp (send VOICE UPDATE to +31620879652) every working day. Or well, almost every working day, because I’m only human ;). Want to know more about me and my work in voice? Check out smartvoicestories.com.

Source: Artificial Intelligence on Medium

(Visited 4 times, 1 visits today)
Post a Comment