Alexa: How Can I Create a Voice Experience People Will Use?

Digital Marketing


In 1962, at the Seattle World’s Fair, IBM demoed the ‘Shoebox’—a machine that could understand 16 spoken words in English (10 of them were actually just digits, from ‘0’ through to ‘9’, but at the time, this was genuinely groundbreaking stuff).

Four years later, an early episode of Star Trek aired, featuring a conman trafficking mail-order brides. But the episode was notable for another reason—this was the first time viewers would have seen the crew of the Enterprise talking to its computer. Straight away, the technology played more than a passive role; the computer contradicts the episode’s eponymous villain Harry Mudd during his interrogation, to which he responds “Blast that tin-plated pot!” You can see the full dialogue, including some brilliantly blunt contributions from Computer, here.

In an era long before Apple or Amazon existed, a voice-enabled future was already beginning to loom large in the popular imagination, but for decades, these early experiments and sci-fi fantasies remained exactly that. Computers got exponentially more powerful, the internet changed the world, and mobile technology put all of this in the palm of your hand, but visual interfaces still ruled in this emerging 21st-century ‘screeniverse.’

But just as screen-centric mobile UX is reaching maturity, a new age of interaction is finally upon us. Siri has been softening us up for a while to the idea of speaking to our devices in a conversational way, but the experience remained clunky in many respects. Today, though, voice-assistants and conversational UI are quickly becoming the norm, and these interactions are becoming more seamless and satisfying every day.

So just how pervasive are voice interfaces right now?

40% of adults now use voice search once per day, while voice assistants like Amazon Alexa and Google Assistant are growing fast, vying for dominance of this fast-developing space. 50% of US homes are expected to have voice assistants within five years. Some say the future is voice-activated. I’d argue that future is already here.

In this blog, I’ll share what consumers think about voice assistants right now from our own 2018 CX trends research, how you can get started with voice, and what that means for marketers.

Our Recent Research Findings on Voice Assistants

We recently carried out some research around 2018’s top digital customer experience trends and found that more customers say they want to use voice assistants this year than other key technologies that are changing how we interact with brands.

What Technologies Do Customers Want to Use in 2018?

Interestingly, people are far more comfortable speaking to a virtual voice via a speaker than they are messaging a virtual participant in an online chat.

This may be because voice assistants are able to remove friction from some everyday interactions that didn’t involve another human in the first place (such as checking the weather), whilst chatbots are often used for customer service and communications, where people still value human contact. Despite the healthy appetite amongst consumers for using voice assistants in 2018, nearly one in 10 people said ‘anxious’ was the word that best describes how they feel about this technology.

Finally, we found that the biggest thing consumers wanted from digital in 2018 was improvements to the products and services they already use, rather than flat-out innovation.

Improvements Innovation Disconnection Integration

For those thinking about developing a voice experience (e.g., an Alexa Skill), the takeaway is clear—don’t do it just for the sake of doing something new; think about how you can use voice to remove friction for your users and help them accomplish the things they’re already focused on.

Perhaps surprisingly, the proportion of people looking for disconnection (i.e., less technology in their lives) actually dropped compared with our 2017 findings—this adds to the evidence that people will be increasingly open to bringing devices like the Amazon Echo or Google Home into their living spaces over the coming years.

Experimenting With Voice

Our first public experiment with voice UIs as an agency was an Alexa Skill for The Higher Lower Game—the App Store-topping game we developed based on Google search data. This was a perfect entry into the world of voice for a few reasons:

  • The game was already successful, so we had an established audience to tap into
  • The simplicity of the game’s mechanic was ideal for a voice-based experience
  • The game is our own IP, so we didn’t have the pressures we might have on a client project

Here’s a quick video of the Skill in use, complete with some top trash-talking from Alexa:

Following that, we got an opportunity to work on an exploratory project for a major client, looking at how we might use Alexa to deliver interactive learning/coaching experiences for young people with under-developed communication skills. We had some fantastic audience research from the client to work with and assembled a small team (developer, UX consultant, and content specialist) to quickly design and build a prototype.

We began with a set of questions around what the technology could or could not do, and, with some rapid prototyping off the back of this, very quickly established that the vision we and the client had been so excited about wasn’t realistic right now. However, we still managed to create a demo that showcased, with a bit of smoke and mirrors, how our imagined interaction might work once the technology has moved on a little. Although we didn’t come out of this with a user-ready product, the project was still valuable:

  • The client was smart enough to treat this as a discovery project, rather wasting thousands of pounds commissioning a full build
  • The project helped all stakeholders understand what the technology can do, which will help inform better briefs in future
  • We’ll be ready to hit the ground running as and when voice assistants evolve

As our Design Director Tom Bradley explains, “We see so many businesses still approaching this type of work with an end product in mind, rather than viewing it as an experiment. It’s hard from an agency perspective too, as an answer of ‘not yet’ can feel like a failure; but sometimes the right thing to do next is to do nothing at all.”

The key is to test and learn quickly so you can pull back when necessary, without missing opportunities to get a competitive edge.

More recently, we’ve been involved in a developing a hotel-booking experience for Alexa, which came with its own set of challenges around use cases, scope, and API integrations. Once again, taking an MVP approach and avoiding a big-bang release has allowed us and the client to test the technology in a low-risk way, without large-scale investment.

After designing an initial conversation flow with nothing more than post-its on a wall, we were able to begin testing the basic UI with people using a role-play approach (being Alexa is always quite good fun!)  This allowed us to refine the flows and the copy before we went anywhere near Amazon’s Developer Portal to begin building it for real.

Whilst this was a very effective way of working internally, it didn’t give us the best set of outputs for sharing progress with the client at the early stages. If you’re in a similar situation, I’d suggest getting the client involved in the project as early as possible and, ideally, have them working with you in your space to simplify communication and ensure the brief and objectives remain clear.

Ultimately, though, we were able to deliver successfully against the brief, thanks to:

  • A small, multi-disciplinary team
  • Low-fi prototyping and testing using role-play techniques
  • A client who understood the value of an MVP approach

If you’re interested in getting started with designing for voice, check out Amazon’s guide here and Google’s here. I’d also recommend reading The UX of Voice: The Invisible Interface.

Looking Ahead

So what’s next for voice experiences?

According to Gadgets 360, the big trend at CES 2018 was “a voice assistant in every corner of your house.” Whilst there continues to a be a big focus on the rivalry between Amazon and Google, we can expect to see voice interfaces breaking out beyond smart speakers into TVs, PCs and other home appliances. How well all of these devices can integrate is unclear—while Amazon and Microsoft have taken steps to get their voice assistants playing nicely together, Google and Amazon are showing no signs of wanting to cooperate.

Beyond the home, voice will also become increasingly central to in-car interfaces, perhaps even before manufacturers ever figure out how to do touchscreen UI properly. Major brands like BMW and Toyota will be integrating Alexa into their cars this year, whilst Hyundai have linked up with Google Assistant. In the car UI space, there’s basically a three-way fight playing out between Apple, Amazon, and Google. Again, it’s a messy situation for consumers to make sense of right now, but the overall trend is clear: voice is the new paradigm.

My own background is actually in SEO, and I’ve barely touched on the implications of voice for search here—that’s something for another blog—but what I would say is that if Alexa manages to monopolize the voice space, that could have serious implications for Google’s search dominance. Whilst the outcome of that particular battle is out of our hands, you can and should be optimizing your content for voice search right now.

5 Key Takeaways

If you want to create a voice experience that people actually want to use, remember these five things:

  • This is a fast-moving space. Keep up-to-speed with the latest research, or, even better, do your own.
  • Start learning to experiment as soon as you can, ideally on an internal project that doesn’t need to deliver a return.
  • Approach voice projects with a lean mindset—create a small, cross-disciplinary team with UX, content, and build skills, and test and prototype fast. Role-playing is your friend.
  • If you’re agency-side, talk to your clients about discovery projects, where it’s ok for the outcome to be “do nothing yet.”
  • Remember that while ditching the screen is a big deal, many of the same basic UX principles apply. If your voice experience isn’t making someone’s life easier, it probably shouldn’t exist.

Have you considered incorporating voice experiences into your marketing plan? What potential do you see for voice assistants for marketers?