Texting, for Better or Worse
The success of Talkspace in the new field of behavioral therapy chat created a hype which my employer wanted to follow. At the time, our telehealth company provided an online consultation platform based on phone and video calls and had established a nation-wide network of regionally licensed behavioral therapists. Therapy chat appeared to be within reach.
I was asked to study these services and evaluate what it would take to enter this new market.
Harder to write than to talk
Therapy chat’s low subscription prices were appealing to consumers because they were cheaper than regular therapy sessions and appeared to provide unlimited access to a licensed therapist.
But unlimited access and the written medium --harder to use to talk about feelings than just talkin--raised our concerns about sustainability and a viable provider’s compensation model.
These assumptions were strengthened by early feedback obtained from therapists who had worked with their clients using messaging and fund the medium to be the most work intensive way to deliver behavioral treatment.
poor efficiency, but very engaging
I built a team of volunteers to test Talkspace and Better Help products. Our objective was to evaluate users’ experience and deduce providers’ workload. We had three weeks to sincerely engage with messaging therapy.
The interactions with therapists were highly rated by our testers. However reproducing “talk therapy” via the messaging medium proved difficult and time-consuming. But paradoxically it engaged patients with dealing with their issues and pushed them to pursue their treatment. The usability issue related to writing created a strong incentive to purchase add-on services and plan upgrades for synchronous interactions like live chat or video conferencing.
Creating a sustainable therapy chat service wasn’t just about building a messaging app for patient and therapists. Rather, messaging was a device for user acquisition and channeling patients towards more conventional means of treatment.
unsustainable as such
Recruiting and retaining providers was a central concern considering the initial resistance we faced among our own provider network. We needed to understand further the inadequacies of the medium with therapy work and to do so we used a standard messaging application to simulate a work interaction between a provider and multiple clients.
Using messaging was 3 times slower than talk therapy to reach an equivalent outcome. Therapy via messaging as a standalone product appeared economically unsustainable in a US economic and regulatory environment. Messaging could have value if providers used it as a tool for patient acquisition and engagement. It needed to be integrated seamlessly with audio, video or in-person therapy sessions. It simply couldn't replace live sessions because no payers (insurance) would recognize the medium as a valid mean of therapy delivery.
The medium had more potential with less regulated care sectors such as some type of counseling, coaching or health education.
a 'workable' messaging app
Regardless of the product viability our simulation helped us identifying a range of care professionals' needs that a standard messaging application wouldn't address. For instance:
- Providers needed to identify the stage of their relationship with a client so they could provide the appropriate attention: prospective clients, clients in treatment, returning clients.
- Messages from a client needed to be aggregated and prioritized in a way that fitted the working patterns of a therapist. We planned to design a sorting algorithm that could organize a client queue according to urgency, velocity but also compliance to treatment and patients usage patterns.
- Clients needed to be identified quickly without reading previous messages. We suggested to design a quick profile view that described the patient through keywords.
An app for writing, not for texting
Patients were also facing a range of usability issues.
- For instance a typical patient's message differed from the typical text message: it was considerably longer, it also tended to be formulated with more care and used punctuation. The communication patterns used during therapy were closer to writing a post on Cora than chatting on Whatsapp. So we suggested to develop a UI that evoked message board behaviors: providing a larger space text entry field that for a typical text message.
- In a typical exchange between patient and therapist the therapist typically leads the conversation with open questions. We suggest to materialize such questions to help streamline the initial exchanges between parties.
- Patients were also likely to proofread and edit their message before sending it. We suggested a "no-regret" feature that prevented accidental submission.
Unlike the companies we were going after we had a large user base, a large provider network, and a proven phone and video infrastructure. For us venturing in messaging therapy was about connecting the dots. But - as it turned out - our then aging user database structure prevented us from tapping into our user base - we only had access to a narrow segment of users and consequently we had lost our head start. We simply didn't have the resources necessary to develop the service from the ground up. The project was eventually parked.
But our work wasn't wasted: it later became valuable for a project about hospital visits management where text messaging became an important feature.