As part of our preparation for Amazon’s Alexa Cup, we wanted to identify the emerging needs of a new audience for the Echo product range, specifically by seeing how this new technology could support people as they grow older.
We decided to run a condensed design sprint so that we could take the research we’d gathered and, following a rigorous process, quickly produce a working prototype that we could demo and test.
Voice skills for the elderly have huge potential. There are 12 million people over the age of 65 in the UK (677 million worldwide), with many experiencing motor or visual impairments which make using screens difficult or impossible.
We conducted a thorough research review to pull key insights from work that had been completed in this space that we could build upon.
“…one of the great triumphs of the previous century – looks set to be one of the great challenges of this one”
Design Council – Transform Ageing
– Older people will often hide the things they are having difficulty with from friends and family because they don’t want to become a burden or ‘make a fuss’.
– A positive outlook on ageing has a big impact on actual quality of life.
– There are common ‘transitional events’ that knock people out of a healthy routine.
– Any automation or assistance should be carefully considered, as it could inadvertently displace another beneficial activity (e.g. getting up to change the radio channel, going out to buy a newspaper, etc.).
It’s easy to think of older people as one homogenous group, but they are hugely diverse with many different needs.
Having identified the potential areas of opportunity, we looked for patterns or themes that could be used to further shape our early thinking. The following areas emerged:
– Establishing and maintaining healthy routines
– Cognitive sharpness
– Emotional well-being
– Social connection
– Entertainment and new experiences
To address any one of these areas, ultimately we would need to be able to produce a ‘shining example’ of how voice assistant technology could cut through with this audience. We knew we would have to:
– Find a reason for people to install and then proactively launch the skill
– Initiate repeat engagements
– Ensure each engagement adds value (one that doesn’t will likely be the last)
– Make sure complexity of use (low) was in proportion to value (high)
– Be natural
At this point, we were able to come up with some early ideas to quickly see if any areas of interest were beginning to emerge that were worth further investigation.
We ended the day by working on a value proposition that was underpinned by our research insights and could then be used to further shape and evaluate our thinking.
Here’s what we came up with:
Instead of reaching a crisis point, the “New Day” skill will help older people in transitional situations to remain happily independent, by making life less difficult through a productive and entertaining daily routine.
Day two was all about ideas. We needed to build on the research from the first day and start to describe the new experiences that could be unlocked by these opportunities.
We generated a number of ideas, developed some of the most promising ones, then conducted a design review with the wider Experience Design team here to get an outside perspective on whether our ideas were clear and impactful, and see how they might be improved.
We prioritised the ideas that had a clear link to the core needs we’d seen in research, to ensure we were delivering a long-term benefit, but also everyday usefulness or entertainment value so that the skill provides some instant gratification.
Our ideas broadly fell into two groups…
– Daily routine for physical and mental health
– Social connection and access to services
We chose to prioritise ideas relating to daily routine, as these had the more obvious impact but also because we felt that using Alexa to provide social connections might actually reduce a person’s opportunity to do this themselves.
The final idea that we chose to take forward was a simple word game that older people could use to get some entertainment each day, but also monitor their mental acuity over time. We felt this could be one example of a suite of skills that could help older people stay mentally healthy, but also physically healthy by managing their nutrition and mobility too.
We knew that to design this skill well, we’d need to make sure it was uniquely-tailored for the intended audience, and by doing this we would create something that could be more easily used by everyone. We also wanted to make sure that a voice product was better than the available alternatives (TV game shows, quiz books etc.)
So, we developed the following principles and embedded them into the design:
– Not overly reliant on technology and no need to be tech-savvy to get started
– Low-pressure and can be completed at your own pace
– Follows a predictable schedule that helps establish a routine
– Challenging enough to create a sense of achievement, but not overly competitive
– Offers a sense of companionship that you don’t get from other word games
The final exercise of the day was to begin to map out and script the experience so that we could start producing a prototype on day three.
This allowed us to identify the obvious things like the ‘wake word’ that people use to initiate the skill, but also the unexpected things that people might run into that could cause problems (e.g. asking the person if they have a pen before describing the letters they need to complete the challenge).
Time spent planning is never wasted. That’s true of lots of things, but especially so when you draft in a developer to build something in a very short space of time!
Whilst our hero developer was busy getting to grips with an unfamiliar technology, the rest of the team concentrated on mapping out the conversation between Alexa and the user, and figuring out our story for the pitch the next day at Amazon’s Shoreditch HQ.
We’d tried to immerse ourselves as much as possible into the world of older people, drawing on research, stories and imagining a day in the life of our user.
This meant that even small decisions needed to be thought through and debated carefully. Things like:
– How quickly should Alexa talk? (we’d seen lots of complaints about reading speed)
– How much information should we give at once?
– How do we make sure it’s challenging yet doesn’t exclude people?
– How does Alexa react when it doesn’t understand?
– What tone of voice should we use?
You’re never going to get all of these decisions right, but it’s important to think of them from the outset.
Once we had the basic flow in place, our developer could start to build the skill for real. It was going to be tight, but we had plenty of momentum by this point.
Along with testing around the office during the day, that evening we had a chance to test the skill out with a representative user for real. We took an Alexa device into the home of an older person (let’s call them Tony) who was a perfect fit for the audience we were trying to reach.
He was independent but living alone, and was facing recent difficulties in his life that could potentially be aided by voice. For example, because his eyesight had recently started to fail, he could no longer drive and needed to rely on buses and trams. Voice could help him get information about his journey as his tablet and phone, with their reliance on eyesight and dexterity, were becoming a source of frustration for him.
As expected, seeing the skill being used in-situ was revealing:
– Tony could immediately see the value of voice: “It would mean I wouldn’t need to bother with programming my tablet”.
– Tony went on to list 20 things he thought he would use voice for.
– Choice isn’t always a good thing: “When Google gives me 10 million things I think I shan’t bother”.
– Setting up the device was difficult: “I wouldn’t know where to start”.
– When our skill asked for Tony’s name, it panicked him slightly as he didn’t know if it wanted his first, second or full name. Nobody in the office thought like this.
– We were surprised at how clearly he understood the rules and got into the game side of the skill. This felt really natural. Once he was in, he was in.
– Onward journeys and onboarding are really important, along with being able to fall back gracefully should the tech fail, making it clear that it’s the tech’s fault, not the user’s.
An early start to get down to Amazon’s HQ in London meant that some of the team only had a couple of hours sleep, but we were excited to get down there, and the fact that we would still be scripting and building the skill, and pulling together the presentation, kept the adrenaline flowing throughout the day.
It had become clear from our sprint that:
– Voice has a huge potential for older people
– There are many barriers stopping the adoption of the technology
Our task then was to focus on a voice skill that removed as many of those barriers as possible, making the first skill they used easy, repeatable, entertaining, relatable and above all, a comfortable gateway into the world of voice.
In short, we wanted to make something that gets older people into voice before they ‘need’ to.
We’d use ‘Tony’s’ story to demonstrate this, and explain our approach.
We received a whole load of helpful tips and guidance from the Amazon facilitators which was invaluable, especially on the tech side.
With 20 minutes to go before the pitch was due, we were still testing the prototype skill and it was finally launched, ready to go, with just minutes to spare for the live demo.
A long story short, we didn’t win. The field was very strong and it was clear that other teams had been working on their voice skills for a while, with some really useful and innovative solutions to everyday needs.
We did get some really positive feedback, however, and we’ll hopefully be launching the skill for real in the near future.
We learnt a whole lot along the way too, and we are in the process of sharing this with other organisations involved in designing voice for older people.
It’s difficult to show a voice skill. So to give you an idea of what we made, here’s the script for the ‘happy path’ where the user goes through the skill with no errors etc.
In the actual skill, we slowed Alexa’s voice down to make it easier for older people to follow along (this is subject to testing) and we’ve designed some screens to go on any devices with screens.
The user’s voice is in bold
Alexa’s voice is regular
“Alexa, open ‘News Letters’
Sonic ident plays
“Welcome to News Letters – the anagram game that keeps you in tune with the news.
The rules are simple – I’ll give you 9 letters every morning, then it’s up to you to see how many words you can make with them, including one special 9-letter word.
Don’t worry – I’ll explain how it works as we go!
Before we start, do you have a pen and paper handy?”
“Ok, I’m going to give you nine letters, one at a time – all you need to do for now is write them down.
“Just say ‘repeat’ if you miss one. ”
“The first letter is A for alpha”
“The second letter is H for hotel”
“The third letter is C for Charlie”
“The fourth letter is P for Papa”
“The fifth letter is A for Alpha”
“The sixth letter is T for Tango”
“The seventh letter is E for Echo”
“Nearly there… The eighth letter is R for Romeo”
“And finally, the last letter is U for Uniform.
”Did you get all of those?”
“Great! Now here’s a clue for today’s nine letter word.
On this week in 1944, the Normandy landings took place – a turning point in Word War Two.
The word you’re looking for is loosely linked to this event
OK, good stuff. There are also over 250 shorter words you can make from these letters – these don’t necessarily have anything to do with the Normandy landings, but maybe you’ll spot a few coincidental connections!
After 3pm this afternoon, I’ll be ready to reveal the special 9-letter word, and see how many others you found.
Does that all make sense?
“Great. Would you like me to remind you to check back in later?
“Is 3pm ok?”
”Ok – I’ll remind you at 3pm.
If you’d like another clue at any point before then, just say ‘Alexa, open News-Letters’.
Sonic ident plays
”Hello again! Welcome back to News-Letters.
Did you find today’s Normandy themed 9 letter word?
“Yes I did, Parachute”
“Well done. I hope I’m not making this too easy for you!
13,000 U.S. and 7,000 British paratroopers landed behind enemy lines before the D-Day invasion. And this week, 75 years later, some of these veterans once again parachuted into Normandy, recreating their historic mission.
Apart from ‘parachute’, there were 258 other words you could make with today’s letters. How many did you get?”
“Great work. That’s about a third of all the possible words.
I have one more bonus fact about today’s topic, would you like to hear it?
You might have also found the word ‘chateau’. Following the D-Day landings, there was a fierce battle for the infamous Chateau de la Londe, which was heavily defended by an SS Panzer division. East Yorkshire and Suffolk regiments ultimately took this German stronghold.
Tomorrow’s letters will be available from 9am. Come back anytime after that to see if you can beat today’s score, and if you think you know anyone else that might enjoy this game, why not tell them about it?
Remember, just say ‘I’m ready for today’s News Letters’ when you want to play again.
Bye for now!”
We see this as a small, simple product that can grow over time. Ideas that went into the ‘great idea but we’ve only got two days to build this thing drawer’ were:
– Having Susie Dent as the voice of News Letters
– Writing, maintaining and sharing voice design principles for older adults
– Differing levels of difficulty
– Echo button integration for people with difficulty speaking
– Sharing your results with friends/family
– Sharing your results with strangers
– Ability to practice more anagrams whilst waiting for the answer
– A suite of entertaining, onboarding voice skills
– Donate to a charity for older adults through the skill
– Topical hints and tips (how to stay warm in the winter, where to get help etc)
– Editorial calendar
– Training the social media team of a charity for older adults to update the skill daily
Contact us to find out how we might help you build a better experience for voice devices.
Read more about voice: