Yesterday, I wrote about how voice interaction requires some effort on the part of the user. When designing for ambient voice interaction (or creating the next generation of devices) there are some ways that we can reduce users’ efforts.
There are ways to reduce this after a user has initiated, but we’ll deal with that later. In the meantime, how can we limit the amount of cognitive load needed to initiate an interaction. Some ideas:
Be the initiator. It takes less effort to prompt a user to answer a question than it does for the user to generate and ask a question first. This can be done in a few ways:
- The device can barge in and ask the question directly. “Hey John, would you like to order a medium veggie lover’s pizza from Pizza Hut for 6 PM?” Being specific in the inquiry reduces the back and forth quite a bit.
- Beep beep. Prompt through an audio cue that would require the user to ask a question. Here, you’re massaging the users dopamine receptors to make them curious about what could be prompting them. This can only be done so many times before its disabled.
- Flash flash. Similar to (2) but the notification is done through light indicators.
Be present. Reminding a user of the device’s presence is one way to limit the cognitive load on the user of remembering that he or she has a device within earshot of them. Lighting up slightly when someone has walked into a room is one way to do this. A clicking noise is another method.
Be an inceptor. Similar to being an initiator except that you’re quietly suggesting something the user can ask. This can range from the benign to the more deceptive (the latter should be avoided). The Echo Show displaying a list of things you can ask it is one way of doing this. Releasing a scent of fresh pizza to get the user to want to order pizza is not good practice. In between, you can display on other screens imagery for relaxation or start playing music. This might lead the user to request different music.
Be a room reader. Knowing how many people are present in a room can help any particular user feel less conscious about making a particular command. A device can perhaps change colour to indicate that it knows a particular user is present when there’s a group.
If you replaced the ambient device with a person with very high emotional intelligence, how would they make it easier for someone to talk to them? That’s really what we should be asking to help people start to interact with our devices by voice.