What Would It Take to Provide Evidence-Based Emotional Support at Scale?

What Would It Take is a series of short-form, podcast-style recorded conversations with problem-solving change agents committed to putting science to work for society. 

This is a transcript of episode 2 in our pilot season focused on projects aimed at improving health outcomes. It features Glen Moriarty, CEO of 7 Cups, and SIL's Kate Green Tripp discussing the question What Would It Take to Provide Evidence-Based Emotional Support at Scale?

We invite you to listen to the recorded conversation or read the transcript below.

Episode 2: Transcript

Kate Green Tripp: Welcome to episode two of What Would It Take – a conversation series from Stanford Impact Labs designed to expose and explore what it looks like to tackle social problems with a solutions-focused orientation and a dedication to partnership. I’m your host, Kate Green Tripp.

Today we are joined by Glen Moriarty. Glen is a licensed psychologist and the founder and CEO of 7 Cups, an on-demand emotional health service and online therapy provider that connects real people with real listeners in one-on-one chat. Together, we’ll consider the question, What would it take to provide evidence-based emotional support at scale?

Thank you, Glen, for being with us here today. 

Glen Moriarty: Super happy to be here. 

Kate: I'd love to start the conversation by digging into the problem a bit and asking you to help us understand what the data shows when we think about mental health need, what is often referred to as the mental health crisis.

Glen: Yeah. So I think mental health crisis is a good way of saying it. There's, you could say, like one in five people who need help at any given time.

We can look at the stats on depression, anxiety, drug use, suicide rates, etc. It's all going up and to the right, which is not good. It's increasing. In the U.S., broadly 50 percent of the people that need help can't get it. In the rest of the world, in a lot of places, 80 to 90 percent of people can't get help.

And I think mental health can feel a little abstract as a concept. And so I like to think of another way of it as like human suffering. So mental health can be seen as different expressions of human suffering. And so we have a lot more suffering. You can think about family members you know, friends you know, your kids’ friends, people in your town, city, or state. We just all know a lot more people that are struggling now than we have in the past. And I think that's problematic and not getting better.

Kate: How are folks getting support? What are the different forms of support? I'm curious. Not to be inappropriate about it, but it's almost like an economics question a little bit in terms of supply and demand. So help me understand that a bit. 

Glen: Yeah, it's a huge supply and demand problem. So the primary way people get help is through clinicians. So these can be therapists, social workers, psychologists, psychiatrists. The way we visualize it is you've got just this big bucket of demand, people that need help and that bucket is growing, you know, post-Covid. It's a deeper and deeper bucket. And then you have a limited pool of clinicians.

Ideally, those buckets would be the same, or at least closer (in size). Let's say you have—on the people that need help bucket—let's say that's a ten in terms of size and the clinician bucket is like maybe a one. And so the level of people on the side of the one, there's no way there's going to be enough clinicians to meet the demand of the side of the ten.

And so there's just this enormous mismatch. And, you know, it's pretty pricey. You know, your average therapy, it's $150 an hour cash. So the prices are pretty high.

And then you have thankfully things like support groups, right, where people can meet and provide care to one another. You have the innovation of AA that, you know, Bill W. (Bill Wilson) launched decades ago that helps people with substance use challenges. And those things are, by and large, free right and they’re beautiful.

Kate: So given the supply and demand problem and the larger backdrop of just how significant the need is, I'd love to hear more about 7 Cups. What is this? How does it work? Who does it serve? 

Glen: Yeah. So the story on 7 Cups is, I'm a psychologist by training. I'm also an entrepreneur. So my first company was a learning company. So when I was a kid, learning was offline. Then it moved online. Dating was offline. Now it's of course moved online. And so I was struggling with what is the behavior that's offline that hasn't yet moved online. And my wife was listening to me (she's a therapist) and it occurred to me the need to talk to somebody is the behavior that's offline that hasn't yet moved online. And so a lot of people, they just need somebody to talk to them. 

And so the idea is, is that anybody, no matter where you're from, no matter what language you speak, no matter how much money you have, no matter what resources you have, you can go online and you can talk to a compassionate listener. Somebody who will provide active listening support to you. And that's like summarizing, paraphrasing, empathizing, just helping you feel heard and understood.

And, you know, the gist of it is like the listener, it's their job to imagine themselves as you and say, like, okay, if I were you, what would I be thinking and feeling? And you just say that back to the person in your own words. And that feels just amazing. Whenever you've heard somebody, you've had the experience of somebody really listening to you, it just feels great, it’s healing.

It's helpful, it's therapeutic. And that is about 50 percent of what makes therapy work. So they call that the therapeutic alliance, which basically means the therapist likes you. They're on your team and they want to help you deal with or face the challenges that you face. And so we started 7 Cups and went through Y Combinator, which is right near you.

I was the first listener. We had a bunch of listeners from all up and down the universities along the California coast. And then somebody who was working at NAMI (National Alliance on Mental Health) said, Why don't you put a little banner on the top of the chat that says, One of the best ways to help yourself is to help somebody else? And all through the first several months of this, the demand was so high, we couldn't get enough listeners.

And then when we put that banner up, we started converting. So some percent of the people that were getting help signed up to give help. And so now we have 564,000 listeners, 189 countries, 140 languages. And we have 80 therapists on the platform as well. And then we have forums and group support chat rooms and treatment plans or growth paths.

So there's a whole robust kind of support or treatment offerings on the site. But the core of it is just the inherent human need to talk to somebody when you're going through a hard time and then have somebody really listen to you.

Kate: So it's a place where you can either seek the support of a listener or become a listener yourself. And it sounds like I'm hearing you say in some cases folks are doing both. 

Glen: Yes, that's right. 

Kate: So, listening is not the same as therapy, but it is a crucial element of why and how therapy works. Someone volunteering to listen to another as I understand it, that modality is often called peer support. Can you say more about that?

Glen: Yeah, peer support is largely something called like lived experience. So for example, if I lost my mother to cancer, I would have already gone through that process.

So I knew what that's like. So then somebody else who's in the process of losing their mother to cancer, I would probably have a lot of insights and ability for them. So really basic things like, Hey, you're not alone. I have all of this assumptive framework, like all the things that you're already going through. I already know about.

I've already kind of crossed these bridges. Or I'm a vet , you know, and I've seen active combat and I know what it's like to come home and, like, be dreaming about that. And now I can help you do that, just coming home with that process. And so that's multiplied across almost every domain in life. And the people that have that lived experience, they want to help others because it redeems their experience and gives them meaning.

So for us, you can search for listeners by topic issue, you can search by language, you can search by country. And the more of those things you hit, the more potential there is for there to be meaningful peer support. 

Kate: So it’s by harnessing the kind of willingness and, it sounds like, the power of a peer support base, you're both serving the desire to help, of the peer supporters themselves, as well as the need. Is that correct? 

Glen: Yeah, that's exactly right. 

Kate: You clearly then have the infrastructure in place to support folks seeking, listening and somehow being matched with listeners. How are listeners supported or maybe better said, how are listeners trained to kind of hold that role for others.

Glen: Yeah, so great question. So our listeners go through an initial training program. It's an hour-long course. They have to watch videos, do exercises, pass a test. Then we put them in front of a bot that acts like a student who's depressed. They have to make that bot feel better. Then they make an oath to treat people with dignity and respect and then they get enrolled in our continuing education program. And we have our own academy. It's got five different programs in 70 different courses. And all of this is free. So anybody can sign up and kind of go through it. And then once they become a listener, we also have something called Just-in-Time training, which is AI support.

When I was training to be a psychologist, my supervisor would sit behind a one-way mirror and I would have a bug in my ear, like a little speaker. And I would be interviewing the client and she would say, Ask about substance use or Ask about this relationship or, you know, Go back there, That's important, Say more. 

And so with her kind of guiding me in the supervision process, I would, you know, be interacting with the client. And so what we do is we've modeled our Just-in-Time training after this so that if you're if I was chatting with you and I was stuck, I could press a button and then the generative A.I. would create a response that I couldn't copy and paste, but I could be inspired to use. 

We've published some interesting research on that, how that's related to helping listeners feel more supported. So it's both kind of traditional coursework or practical work married with the Just-in-Time training. And then in addition to that, there's peer supporters for the listeners. So any time you're going through a hard time, you can chat with another listener or more advanced listener to support you.

There's a listener support room that's happening in real time as well. And then there's mentors as well. 

Kate: So it sounds like the AI component is modeled on sort of this historic established best practice that, you know, a traditional clinician or therapist would go through that we often think of as the supervision process, getting guidance, informed guidance. So how is that AI tool being trained?

Glen: So the AI tool itself has been trained – you train it on transcripts. So we created a bunch of our own transcripts, like thousands of conversations. And then we found best practices transcripts in all kinds of domains, different active listening, empathizing, basic counseling skills, and reviewed talks with different experts.

Kate: This all happens through chat. Is that correct? 

Glen: That's right. It all happens through chat. And the thing is, listening is not always easy. Like I'm sure you've remembered in your life when somebody like a friend says something and you're like, What do I say? 

Oftentimes as a listener, you have a big heart and you really care and you want to communicate, I care for you and I'm listening, but you can get self-conscious. You can get stuck and this is a little almost like phone-a-friend, press a button and pop, and it just becomes easy to say, Wow, it sounds like that was really distressing or whatever. And so eventually, their confidence as a listener grows. So basically what we're seeing—with almost like all these AI elements—is they become a copilot for everything.

So if you're a programmer or a junior programmer and you're just stuck, you can't problem-solve, right? You use the AI co-pilot, it generates the code for you, you get a little bit smarter, you get a little bit more confident. And so it's the same idea with listening. Over time, folks don't really use it really all that much, but it helps them kind of initially like a set of training wheels (or floaties if you're learning how to swim) and then once you can swim or once you know how to ride your bike, you're good—but it's just helpful so you don't have to be scared.

Kate: Help me understand how research plays a role here. 

Glen: So we're working with Diyi Yang and she's doing a sort of training program for our listeners. So, imagine. Remember I described how the listeners go through the initial course and then they chat with a bot that acts like a student that's depressed and they have to make that bot feel better? The thing about her models is like light years ahead of the very kind of basic models that we have. So much smarter, much more advanced, much more specific in terms of actually really helping the listeners develop better skills, especially in the motivational interviewing department, which is basically a way of providing support to people based on the level of change where they are. And so she's done lots and lots of research in this space. So the training is basically a mechanism to help listeners do a really good job listening.

Kate: So Glen, you’ve shared how 7 Cups as a platform works, and how research and machine learning play a role in helping you connect members and listeners. And we’ve also touched on the vastness of the need here when it comes to mental health. So I’m curious: what does it take, do you think, to provide evidence-based emotional support at scale?

Glen: Yes. So we designed it to be free. We wanted to try to really build a free behavioral health system. And that was the free emotional support system. That was the idea.

We started realizing, we can refer some small percent of people to therapists on the platform. And so then that helped us generate some income. And then we started selling branded versions of the platform to like employer groups or health systems, so 99.8% of what we do is at no cost and point two percent is charged, and that keeps the lights on for the whole operation.

And in terms of the scale, a helpful way to think about this is like a triangle. So at the lowest base of the triangle, so like the biggest part of the bottom part of the triangle, you can imagine there being like prevention efforts. And then you go a little bit higher up and you've got listening.

So people for whom the prevention stuff alone is not working, they just need somebody to really listen to them. So that would be the next level up. Then you might have peers—these are non-licensed people that have been trained in specific ways. CETA (Common Elements Treatment Approach) is an example of this—behavioral health specialists or peer supporters.

And then you go up higher and you get your clinicians and your more specialized folks. And so the problem now is that we're sending everybody to the very top part of that triangle. You know, if you think about it systemically, like globally or nationwide. And so obviously they're all clogged up because, as we talked about, the supply and demand.

What we need is a way to triage folks right away and say, hey, look, it's likely that you're going to be really helped by just prevention and listening. And so rather than you going up here, just come down here. For this type of challenge, you can be really helped by behavioral health specialists. These are people with more skills, but they're very affordable. They're not going to cost as much as a clinician. And so we funnel all those folks there, and then the folks that are really in distress can get referred to the folks at the top, at the top of the triangle, the specialists. And so really the solution, I think, involves scaling up all of those elements.

And of course, we've been working on primarily the listening side of it and the prevention side of it with that workforce of folks and we have a tiny sliver of therapists at the top and we would like to add a behavioral specialist help level too. 

I think we're much more bighearted and generous than we realize and if you can make it easy for people to give to one another, they will. And that's kind of part of what we're proving now. And so as a system, if you let people care for each other and you make it easy, I think you can make a big dent in a lot of these challenges 

Kate: Let’s hope. Glen, thank you for being here today and for talking through this question. 

Glen: Thanks for asking. 

Kate: Thanks for listening to What Would It Take – a conversation series from Stanford Impact Labs. To learn more about Stanford Impact Labs and how we partner with communities to put social science to work for society, please visit impact.stanford.edu