Why voice assistants sound like apps
To come up with something new you have to copy, transform and combine (for more on this check out Everything is a remix). First websites looked like newspapers, and first apps looked like websites.
You can call it a lack of creativity, but using ‘old tricks’ in a new media creates familiarity and helps adoption. Skeuomorphism is a good example of a intermediate step towards abstract interfaces that we are familiar with nowadays.
But can using ‘old tricks’ also create disilusionment with new media? Chatbots and voice assistants are all the hype nowadays. Many of them sound a lot like apps. This is probably a growing pain, as ‘after dominating the fundamentals through copying, it’s possible to create something new through transformation, taking an idea and creating variations’ (still ‘Everything is a remix’).
How do you then make a voice assistant (or any new media) a success? I’ll use booking tickets as an example.
1 It should be faster than another way to do it
Buying a ticket is a task that requires you to double-check your input and compare options. Using voice for it, seems cumbersome and slow. Filling out a form on a website is actually faster than using a voice assistant in this video by KLM:
As Mirabeau points out in their research, the majority of people who are searching for a flight don’t have a set destination and travel dates.
2 It should fit the task
Searching a ticket is so much more that just picking one at random (or so it seems that the assistant behaves).
What about comparing prices on different departure times? How can you notice that travelling on another day is cheaper? Based on what does the assistant choose exactly that departure time?
It looks like KLM did make various dialogs possible. However, there are still only 20 reviews for the KLM assistant after 1 year, from this number I assume this assistant was not a great hit.
Searching for travel itineraries is a highly visual task, and it isnot a great match for a voice assistant. Another example is from FlixBus.
3 It should fit the context
Most of the videos that I can find of the voice assistants are filmed in a calm area where you can sit next to your phone/smart speaker and look for your tickets.
I believe the added value of these assistants is exactly the opposite — that you can ask questions on-the-go, get quick answers. Like when you are running around the house packing and want to check the weather for your trip, or following a recipe and want a reminder on how long to cook the pie in the oven. Finding a ticket requires focus and attention, this is hard to do with a voice assistant, as you have to hold a lot of information in your short-term memory, and, unlike a visual interface, you cannot skip over the less important bits to quickly scan your input
This means that
To make any new media a success you should focus on the good old human-centred design process, and look for solutions to problems that users have, rather than looking problems to you can solve with the new technology. Having a voice assistant is cool, but do your clients actually need it?
Related: read about following trends in design.
