Hi friends,

Headspace has been making some bold moves this year. 

Under their new CEO Tom Pickett, they’ve launched direct-to-consumer therapy, a new stratified care model and Ebb, an AI companion.

I wanted to learn more about all of this, so I sat down for a chat with Jenna Glover, the Chief Clinical Officer at Headspace.

We talked about the “why” behind these moves,  what they’re learning about Ebb and AI companions, how client and employer expectations are changing, finding differentiation in the market, measurement-based care, the potential of wearables, and more.

Let’s get into it.

Steve: Jenna, it’s so great to be having this chat.

Let’s get straight into it. You recently announced some big updates to Headspace’s care model. What I’d love to know about first is the motivation behind this shift. Why are you pursuing this path towards stratified care?

Jenna:  Thanks Steve.

So I've seen a lot of different innovations in the field over the years. But I think two areas where we probably have not seen enough innovation are in how people enter the mental health care system and in assessment.

I think you wrote in one of your articles how we’re all saying, ‘right care, right time’. That’s just the tagline, but at Headspace, what we really want to start doing is being a leader in personalised and dynamic assessment to start the treatment journey. That really is the only way to get right care, right time.

The analogy I like to use is to think of mental health care as baking a cake. You have all these ingredients, like the assessments, the evidence-based care, culturally responsive care, etc.

If you bake a cake and you forget to put in the salt, you can't pour the salt on the top when it comes out. I feel like that's what we've been doing in mental health care. We baked a cake a long time ago, and we've been trying to pour salt on top because we missed an ingredient. That ingredient is thorough and comprehensive assessment of all people.

So at Headspace, we're trying to re-bake the cake. We're not just trying to add a little feature, we're trying to completely redo it.

Steve: And so why the shift from stepped care to stratified care?

Jenna: We have a very strong point of view that has not changed: that the majority of people need subclinical care, and we should offer that.

But in our previous model, people came into the subclinical level, and if they needed some type of support beyond content, they went into coaching. And then maybe they would be identified as needing therapy.

That was a really important signal for us - that even though the majority of our members are being served by content and coaching, there's a large number of people who do need therapy services. And we asked, is there a more efficient way that we could route them into clinical care to start with?

We wanted to work on improving this experience for members, the model and making Headspace their mental health companion. Most people don't know where to start when it comes to mental healthcare, and we wanted to create a better experience for bringing people into the system.

We really wanted to have a point of view of, like, okay, what are you here for? Let's do some assessment. Let's make sure that assessment is dynamic, and then let's land people exactly where they need to be. And so maybe that's straight to therapy, or maybe it stays in subclinical care.

But I think the important part of that is if it stays in subclinical care, that there's a specific plan that's given to you. It's not just like, ‘you don't need therapy’. It's, ‘hey, here's specific content that would be very helpful for you’, or ‘here's how you can work with our AI empathetic agent that's going to be really helpful for you’.

What we aim to be is this companion who is with you. That makes it easy to care for your mind. That makes it easy for you to know exactly what your mental healthcare services need to look like, and that is bespoke to each individual. 

So I think we wanted to take that vision, that most people need subclinical care, and make sure that the subclinical care you get is highly personalised. And if you need clinical care, then you have a direct on-ramp to it.

Steve: OK but a lot of people have wanted to build a care model like this and haven’t succeeded. What’s hard about it? Why hasn’t this happened already?

Jenna: The reason stratified care has never been done at scale is because it’s really expensive to do so.

Prior to Gen AI, it would require a qualified mental health professional who is administering assessments to people. We've had some technology to provide assessments, but you would still have to give a whole barrage of assessments. A bunch of people, like 75%, don't need clinical care, and they probably don't need to take a PHQ-9 and a GAD-7, they need a PHQ-2, and then maybe something about readiness for change.

When I worked at the hospital as a therapist, for example, if you came in and saw me, we would do an assessment, and I'd be like, ‘Oh, I want to do an ADHD assessment. I'm gonna run down the hall and grab that measure, or ‘I need to screen and make sure this person doesn't have bipolar disorder”. And so I could make those decisions in a dynamic way.

The alternative [in the digital world] was to ask people a bazillion questions, where people are dropping off and sometimes never get care, or they're taking unnecessary assessments.

Our point of view is that Gen AI now offers something different.

We have been developing Ebb [Headspace’s AI companion] for a couple of years already, and so we have this opportunity to have a really warm companion come in and say, hey, let me just administer the PHQ-2 in a really friendly way. And then, depending on how you answered, be able to say, I need to now do the PHQ-7 instead. Or because you said something specific, I'm going to add the GAD. Or I'm going to add maybe the SCOFF to do eating disorders. And so every person's assessment now becomes personalised.

It's not just a personalised care plan, it's a personalised assessment.

I just don't think anybody could have afforded to do this at scale before Gen AI came into play. It was just too expensive.

Steve: Why? Because you have to throw so much clinician time behind it to do the assessments?

Jenna: Yeah, that's correct. 

And then just thinking about state licensure, just trying to scale up a network where. Even if you're trying to do nationwide care in the United States, you couldn't even have a centralised intake clinic with just one person to do it. You'd have to have professionals who are trained in every single state to do it. And so it's just so cost-prohibitive, even though I consider it the best kind of care.

📣 THR Pro: A Membership for Mental Health Operators 📣

If you want to join a community of top mental health founders, executives and investors, consider becoming a THR Pro member.

As a THR Pro member, you’ll get access to additional insights and data each month, including the full insights from our most recent research report on How People are Using AI for their Mental Health and our article on What Happened to the Mental Health Startups of 2020.

Steve: I’m quite interested in the measurement-based side of your care model. How do you think about doing measurement-based care with the different populations you serve?

Jenna: Our general approach in our V1 is that everybody should start with the gold standard measures of PHQ and GAD in general assessment. That's how we do our measurement-based care.

But we also think it's really important to assess people's readiness for change, understanding how somebody differs who maybe scores moderate on a PHQ, but who's in pre-contemplation versus action.

The general things we need to screen for at this level are those like anxiety and depression symptoms, and then always-on listening for people who are reporting things about, say, PTSD or eating disorders. Then there’s readiness for change and people's preferences. For example, do people want to work with a human? Do people have certain preferences in terms of male, female, race, or ethnicity?

That's our general view of a population screener.

Steve: And how are you redoing these as somebody engages with their care plan?

Jenna: Right now, if you're within the clinical level, every two weeks, it gets engaged.

On the coaching side, at subclinical, the period of time is longer. But this is one of the things we're looking at, like, what is the right cadence for people to continue to check in and do symptom check-ins?

That's one of the biggest parts, so I'm glad you're asking about it. It's not just about triage and personalisation, it's about adaptation. And so what we need to be doing is every few weeks checking in, and engaging by being like ‘hey, we can see you're not doing your care plan, do we need to refresh it?' or ‘let's look at your symptoms, and if the needle's not moving on those, what could we change in the care plan?’

Steve: Do you expect a lot of adaptation to the care plans from getting this information more regularly?

Jenna: I do. I would say, just from my own experience as a clinician, care dynamics or care journeys are pretty dynamic.

As people move through care, there are different things they want to work on, and you just uncover things as people go through their journey. And so I do think there's a fair level of adaptation that happens in a care plan.

As therapists, we often see what we call “COWS” (Crises of the Week). People can stay in care forever if you only focus on COWS. They're loud, they're noisy, and you spend your whole session working on them.

And so I think there's this really interesting balance that we're going to have to learn our way into, around how to adapt to somebody's evolving needs while also making sure the underlying thing you're working on gets enough attention, so people actually make progress.

Steve: Do you think Ebb is in a better position to deal with some of those COWS? From my own experience, I think those types of crises often crop up at specific moments, moments that could be days or weeks away from someone’s next therapy session.

Jenna: Yeah, I think this is the best use case for Ebb.

What we are seeing in our early signals is that the two top things people talk to Ebb about are, first of all, navigating difficult situations. So, ‘how do I have this conversation with my boss?’ Or, ‘I’m heartbroken about this breakup, how do I move forward?’ This is by far, the number one thing people come in for, which I think is totally a crisis of the week.

The second thing that people come with is a feeling of overwhelm, and that could be being overwhelmed by anger or anxiety or sadness, and a desire to find peace.

It's always interesting to me that coming in with this overwhelm and a need to find peace is a recurring theme. I do think that Ebb is a strong use case here. We're already seeing that people are coming to Ebb in moments of distress and that Ebb can help resolve those things for people.

Steve: Why do you think Ebb is good at that?

Jenna: In general, the research shows that motivational interviewing is just such an effective strategy and practice.

I think it's good because there's something that just seems to be quite healing about when people are able to get emotional validation. It also helps them gain insight. So first of all, they can say ‘I feel understood’. There's something soothing about empathy and that reduces the intensity of an emotion.

And then the second is the open-ended questions and the reframing. That helps people shift. Even when I'm using Ebb myself, sometimes I'm like, ‘damn, that was a good question, I hadn't thought of that’. It frees up something, it unsticks people.

People come to Ebb when they’re stuck. And because Ebb is trained in motivational interviewing, it unsticks you. And I think it's through those two processes of empathy that soothes, and open-ended questions that adjust or shift frames, that people find benefit.

Steve: OK, super interesting. But I’m also interested in the context surrounding this stratified care model. How does it fit with your members’ expectations of mental healthcare today? Have you seen any changes in terms of what they expect from care?

Jenna: I think so.

You get three cohorts of members who come into our system.

The first cohort is those who usually have a mental health history, and when they come in, they're often like, ‘I am here for X and I want X immediately.’ And they used to find our system quite frustrating because it was hard to find X. And so stratified care has been helpful for this cohort because it's clear what the options are, and we will route somebody into the right level of care for them. But also, if you raise your hand and you're like, ‘I know what I need’, you get there. So that's helpful.

The second type of user is somebody who comes in and they're just there for curiosity's sake. They might just be there because somebody recommended it, or it's a benefit.

And then the third cohort, and this is our largest cohort, is people who are in stress but don't know what they need. It's not about a curiosity, it's about an uncertainty. What we’ve found is that when we ask these people if they would like somebody to help guide their journey, they're very interested in that.

In general, I do think there's a large number of people who still don't know. They know their mental health is important, but just don't know what that means or what that looks like. I think that's why we wanted to develop stratified care, so that from the moment someone arrives, we can learn some things about them and then, based on what they're telling us, guide them to the things that are going to be the most helpful for them.

Steve: Let’s talk about some of the other stakeholders in this ecosystem, specifically employers and insurers. Is this new model a reaction to their evolving needs? In what ways?

Jenna: From the employer side, people are very cost-conscious right now, and they want a strong return.

For a long time, it's been like, ‘what is your therapy offering?’ And we've tried to respond to that by building our network and making sure we do have stratified care that can get people straight into clinical. But I think now, we're trying to come to them and talk about the larger return on investment, which is to stop buying a solution and paying all this money for something that 10% of your population uses. We're trying to offer something that the vast majority of their employees could use and get benefit from.

We have higher engagement rates than most of our competitors, and so our hope is to be able to show that we have more people engaging in this model and that people are able to get truly better, faster. We're working on our assessment and ROI studies for stratified care right now.

Honestly, Steve, on the employer side, it's so hard to prove like, ‘hey, we stopped this many people from ever needing therapy’. It's a preventative approach. But I do think when you can show higher rates of engagement across the ecosystem, and that when people are still showing benefit at subclinical levels in terms of goal acquisition or improvements on the perceived stress scale, those are things employers do see as meaningful.

The other thing employers are really interested in is the potential for Ebb to coherently bring together benefits navigation. Employers are offering so many different resources right now. We talk about HR leaders having point solution fatigue, but their employees do too. What we're working towards is a world where Ebb not only is having a conversation, but all of a sudden, if you're talking about, say, fertility problems, that if your employer had a care solution for that, we could surface that resource. I think that employers are really excited about that component, especially at a subclinical level, of knitting together things that are going to help the total cost of care.

Steve: I'm interested in your take on how the ecosystem is evolving more broadly. It seems like a lot of care models are converging. Everyone's been expanding services and trying to build some level of personalisation. How do you think about creating separation versus other players, especially in the employer market?

Jenna: I think people are converging, and I think AI is providing that opportunity.

AI is allowing people to do things at scale they couldn't do before, and I think there are a couple of things happening. We know that personalisation is really important. We know that routing is really important. And it has not been possible to do these things at scale until this technology has been in place. And so now everybody sees the opportunity and is trying to grab it.

I do think there will also be an opportunity to differentiate for whoever breaks apart the fifty-minute therapy session. Answering questions like, ‘how much am I going to need to talk to my therapist about?’ Or ‘how much faster can I get better if I am talking to Ebb or whatever AI agent or AI therapy assistant or AI coach?’

If you’re talking to Ebb about the crisis of the week and practising those skills along the way, I think you need fewer and fewer fifty-minute therapy sessions. And for a subset of people who are in that subclinical group, you might not need therapy at all. And so I am interested to see who differentiates themselves in terms of how they use therapy in the future, their length, and their episodes of care.

Steve: What else are you excited about right now?

Jenna: I think a lot about how measurement-based care will continue to evolve.

For example, just how good could an AI be at screening? Right now, it takes seven years for somebody to manifest eating disorder symptoms and get treatment, and it's the disorder that has the highest mortality rate. I'm excited about an AI that could just be so much better at bringing in a variety of assessments and capturing people's actual symptoms and the care they need.

The other next-gen part of MBC, that I think you've talked about, is this idea of wearables - even just like geolocation on your phone is a good indicator of how depressed you might be - just those inputs that could be collected passively, because the hard part of measurement based care is getting people to fill out a lot of forms.

Completion of measures is so low, and so to say MBC is done to fidelity across digital mental health is probably ingenuous. So, how do we really start doing MBC to fidelity, and how do we do things where we can just take these inputs, these passive inputs and use that to inform care in a meaningful way? And for health plans and payers, if we could show that bio data, that's something where I think you start differentiating yourself in a really unique way as a mental health provider.

Steve: I actually just got my new Whoop today, so I’m very excited to see what new data it gives me.

Let’s switch gears to talk about the launch of Headspace Therapy, your direct-to-consumer therapy offering. D2C therapy is a competitive space. How do you think about positioning Headspace in a way that's differentiated and compelling to a consumer?

Jenna: So there are a couple of things that I think are unique to us.

So one, in this space, we don't require a subscription. You're not buying a month of services, you're not buying a certain set of sessions, you’re signing up for a single session. That way, you can see what the experience is like. You have the session, and then you can decide if you want another. You are being charged by session, and since the vast majority of our sessions for the majority of our members are covered by insurance, the average cost is $35. So it's very affordable, and you're paying for what you're using.

The second is that as soon as you sign up for Headspace and you have that session, you get three months of Headspace content for free. So you've got access to this content library that's award-winning and has five thousand pieces of content. In addition, you have an empathetic companion, Ebb, who you can use to process those crises of the week or who can give you bespoke content recommendations based on what's happening with you. So much of evidence-based care depends on whether you're doing in-between session work - that's how all of those treatments are actually shown to deliver efficacy - and so our unique value proposition is that you come in and we're giving you this robust library where you can do that work in between sessions. We have over sixty-five published studies that are peer-reviewed, that show that using our content and care together, you're going to have better outcomes. Our therapists can prescribe those things to you in addition to you exploring them on your own.

And then we're just a known brand that has warmth and trust.

We want to be your mental health companion, and we want to be that end-to-end. That end-to-end experience is something I think nobody else has. And I just really think we provide you with better value for your money and that you're signing up for something that you're using instead of signing up for a gym membership that maybe you go to and maybe you don't.

Steve: And what about AI therapy?

A lot of our competitors are leaning more into things like AI therapy. While we are an end-to-end company, we know that the biggest end of mental health need is subclinical. And so we really lean into AI in the subclinical system space and how it can bring somebody into care and can support someone's care journey when they're working with a therapist.

I think that's a unique bet we are making - that there's a real value in developing that subclinical companion that can complement any aspect of that end-to-end care, rather than going hard into AI therapy and replacing a therapist. We think you can still complement therapy and change the way therapy is done. But I think that's a unique thing that we're leaning into that might be different to other players.

Steve: Jenna, I have to say, I’ve really enjoyed this catch-up. Thanks so much.

Jenna: Thanks Steve, and thanks for all you're doing. It was great to chat.

That’s all for this week. As always, I hope you found it insightful. Shoot me an email to let me know what you thought.

Keep fighting the good fight!

Steve

Founder of The Hemingway Group

P.S. Feel free to connect with me on LinkedIn

P.P.S. If you want to become a THR Pro member, you can learn more here.