In Dialogue—With Our Devices

How conversation design can make technology more accessible, and people more empowered

As technology evolves, making digital experiences feel more human becomes more and more essential. After all, innovative new interfaces can only help those who can access and use them. That’s why the emerging field of conversation design is so promising: It lets people interact with technology by simply talking to it. Consider how we retrieve information. Not so long ago, we’d open an airline app to check on a flight, or trawl through food blogs to find a top-rated restaurant. Now we simply ask that question aloud to a digital assistant, or type it to a chatbot.

Eventually, we can expect to have conversations with any device, service, or platform. This type of interaction feels natural for us; after all, human speech goes back 150,000 years. Computers, which operate according to strings of code—not millennia of linguistics—have a lot of catching up to do. That’s where conversation design comes in: This new field applies the intricacies of human speech patterns to the limitations of technology, finding ways for computers to join the dialogue. With a standard app or website, the inputs are clear: A person clicks a link, taps a button, or swipes left. Think of a yes-or-no interaction: Clicking on a “yes” or “no” button is straightforward. But conversations are more amorphous. Out loud, you might say “sure,” “definitely,” “maybe,” “heck no,” and so on. Language is full of ambiguities, and it’s the job of conversation designers to make computers fluent in each and every nuance.

A better user experience is born when technology communicates like people, and not the other way around. This allows for convenience and frictionless productivity hacks, but more importantly, conversational interfaces also give people new kinds of independence.

We see that happen in ways big and small. For the 62 million people in the United States with impaired mobility, conversational interfaces can help restore individual agency. Someone who once relied on a caregiver to constantly adjust the bed, fan, and lights, for example, can now use technology like voice assistants and smart home devices for instantaneous help (you can see this technology in use, by watching a video on voice-activated room design by my colleagues at Google Home). To make that possible, conversation designers are always considering the most natural ways to communicate essential needs. Someone might say, “Hey, turn on the light,” but likely not, “Please adjust the lightbulb to 100 percent.” That thinking also extends to how devices should respond back to those commands—for instance, should a light say, “Okay,” even if it has visibly switched on?

The work of mapping human language to digital interfaces translates to myriad other needs. Someone with impaired hearing can use apps that transcribe voice inputs on a screen, in real time. Imagine being able to go anywhere—from the coffee shop to the doctor’s office—equipped with a brand new way to talk to those around you. People with impaired vision can check the time, send messages, or confirm the hours of a grocery store without looking at a screen, by asking a voice assistant. And for those with memory loss, conversational systems can provide much-needed reassurance throughout the day. From simple reminders (asking a voice assistant for a reminder that you put the car keys in your jacket pocket) to answering simple questions posed by someone with dementia, conversational experiences can fill in the inevitable gaps left by caretakers, and offer family members relief in the process.

All of these advances make people more independent, but they can also make them feel more connected. Many of these interactions don’t require staring silently at a screen. Other interactions simply enable communication where it wasn’t intuitive or possible before. In one pilot program in which residents of a senior living facility started using voice assistants, participants reported feeling 70 percent more connected with family than before, because of an easier ability to send and receive messages.

With an eye towards these powerful behavioral shifts, our conversation design team has published a guide to best practices. We’re also exploring linguistics further by studying speech patterns—like whether people better comprehend information in groups of three—and collaborating on new features like interpreter mode, which can translate voice inputs into 26 (and counting) languages. There’s a lot at play behind a seamless a conversational interface; designers think about every interaction from start to finish, starting with the question of whether a conversation even makes sense—would someone be better off simply tapping in an app? If not, how can the conversation that follows feel intuitive and empowering—for all users?

Designing for inclusivity—for a wide range of use cases—makes the entire field of conversation design stronger. The more seamless and fluid each interaction becomes, the easier it is for more people to use them, broadening the scope of who is empowered, and when.