Nir Eyal is an angel investor, guest lecturer at Stanford, and author of the bestseller Hooked: How to build habit-forming products.
But today I caught up with Nir to talk about his more recent ventures into conversational interfaces and assistant-as-app.
EJ: You’ve coined the term “assistant-as-app.” Can you tell us about this and how it relates to conversational interfaces?
NIR: I would say that the conversational user interface is the big picture here. Think of it as having three levels of human interaction:
- 100% artificial intelligence,
- 100% human-to-human interaction, and
- somewhere in between.
The assistant-as-app is somewhere in between, it’s a bot-assisted human. You don’t directly interact with the bot, you interact with a human who is interacting with a bot. And I think it’s a more powerful idea because it’s something we already do made easier.
EJ: Can you give an example of how a business might put this to use?
NIR: So, imagine you’re a customer service rep handling customer queries. It would be much easier if the customer texted the company, and their request first went to a bot to narrow down the possible responses. Kind of like IBM’s Watson did on Jeopardy by figuring out the probability of what the likely answer should be.
Then the bot says to service rep, “Here are the three most likely responses the customer wants.” And the human being on the other end decides, “Yep, it’s number one.” That’s something that almost every business today can use.
“We’re not looking for one user interface to swallow up all the rest. The goal is to find the right interface to make a user’s life easier.”
EJ: What are some other situations where assistant-as-app services can excel?
NIR: One of my favorite examples are products that require a dashboard. In the enterprise software world, people are drowning in dashboards. And there are no standards, right? The Salesforce dashboard is completely different than the Google Analytics dashboard which is completely different from the PeopleSoft dashboard. And that means that users are constantly relearning different buttons and configurations. It’s causing what I call “dashboard fatigue.”
Now, imagine instead that you interacted with software through a conversational UI. Take Google Analytics for example, where the vast majority of people use only a fraction of what Analytics can do.
Imagine the service reached out to you everyday with a text message and said, “Hey, you just had a spike in traffic, here’s what you should do next.” Or if they showed anomalies or down time. Analytics can help with these things, but it requires so much cognitive load to figure out, especially for users who interact with these products infrequently. So I think that’s an amazing opportunity for conversational UIs: anywhere there’s a complex dashboard.
EJ: This reminds me of your California Roll Rule. Can you explain what this is and how it applies to business?
NIR: So, the California Roll Rule says that people don’t want something truly new, they want something familiar done differently. Unfamiliar products and interfaces that are difficult to use can impede adoption. Now, this goes back to the conversational interface. We’re already familiar with SMS, we’re already familiar with chat interfaces. So we should use that familiarity to help people do what they want.
Engineers often shout “Look at all the amazing things our product can do!” But what they don’t realize is that if it’s too complicated, if it’s too new–even if it’s amazing–people won’t use it. Because it’s too far outside their mental models, so it takes too much effort to figure out.
“People don’t want something truly new, they want something familiar done differently.”
Apple understood this. My mom never used a computer before the iPad came out, because it made the computer so much simpler to use. I think we’re going to see the same thing with enterprise and consumer applications when it comes to conversational UIs, because it’s something that people are already familiar with.
EJ: Where do you see all this going? What can we expect in the next 5 to 10 years?
NIR: Right now we’re in a kind of Cambrian explosion of interface experimentation. CUIs are still in their infancy. But one of the most exciting branches right now is the voice user interface, the VUI. So if you think about Siri, Amazon Alexa, Google Home, and Microsoft Cortana, these are big companies making very big bets on voice technology. Which is just another version of a conversational UI. In the next five years, we’ll interact with more and more products through voice.
Now remember that not that everything has to be done this way. We’re not looking for one user interface to swallow up all the rest. The goal is to find the right interface to make a user’s life easier.
EJ: So, let’s sidestep over to your book, Hooked. Where do conversational interfaces and habit-forming products come together?
NIR: Well, in the Hooked model there’s a Trigger, Action, Reward, and Investment. The second step is the action stage, which is all about the simplest behavior done in anticipation of a reward.
So, if the conversational UI makes it easier to get what I want, then that’s going to be very helpful in forming these habits.
And that’s why I’m so excited about these conversational UIs. They make the desired goal easier. And that’s where they fit in to the Hooked model. This will make products that use this type of interface much more habit-forming.
EJ: I’m curious, have you got much heat or backlash from people criticizing Hooked for trying to control people’s minds, or get people hooked on technology?
NIR: You know, not when the book first came out. But more recently I’ve gotten some heat from people who haven’t actually read the book.
Here’s the thing, habits are just an impulse to do a behavior with little or no conscious thought. If you want to build or break a habit you need to understand how habits work. One of the benefits of the Hooked model is it explains how products do this, which is also the first step to getting unhooked.
So, how you use that knowledge is up to you. You can use it to build healthy habits using these technologies, or you can use it to break unhealthy habits in your life.
EJ: Have you seen as much interest in the unhooked side of the equation as you have with the hooked?
NIR: Yeah, it’s becoming a topic of national, of international conversation. People realize that we can’t escape technology. Now, the problem is when people point their finger at the technology and say, “Oh, it’s Facebook that’s addicting me.” And that’s just not true. All of us see food, but we’re not food addicts. We all have sex, but most people aren’t sex addicts. It’s the same thing with tech. Many people use Facebook and WhatsApp, but very very few people get addicted.
The technology will do whatever you are predisposed to do with it, so if you want to use that technology to escape reality, there are plenty of people out there who will happily take your money to help you.
So it’s important to keep the problem in perspective. The worst thing we can believe is that these tools are so powerful that we can’t resist them. That’s complete rubbish. And counterproductive. Because it makes us more likely to give in.
Did you know that conversational interfaces are invading articles too?