If you’ve followed our conversation design process, then you’ll have started by designing for a screenless device like Google Home. This is because getting the conversational flow right is easier if everything is in one place—the spoken prompts.

You’ll start creating visuals as you expand your design to include other devices, like mobile phones.

Visuals are meant to supplement the conversation and therefore are not required for every turn in the dialog.

Think about what information could be better represented on a screen. Ask questions like:

• Would it help users to see an image of what’s being discussed?
• Would it be easier for the user to make a choice from a visual list rather than a spoken one?
• Is there additional content that, though not critical for the conversation, would be helpful to show users?

Obviously, visuals will only be displayed on devices with screen output. They’ll automatically be resized and rearranged to fit the size of the screen on which they appear.

Visuals will never appear alone. They’ll always be accompanied by a spoken prompt and/or display prompt. They should also be accompanied by chips.

Think of these visual components as templates with required and optional fields. All you need to do is specify the data (i.e., the text strings, image files, etc.) that you want to appear. See the “Requirements” sections under each visual component for details on what you need to specify.