Deconstructing the Practical Value Proposition in Digital Therapeutics

As stated in my previous blogpost “Digital Therapeutics: a bumpy road to commercialisation”, the Practical Value Proposition (hereafter PVP) is the backbone of successful commercialisation for Digital Therapeutics (hereafter DTx) companies. However, oftentimes DTx solutions insufficiently prioritise the PVP over the theoretical value proposition (hereafter TVP) — this being the main reason for failure in this segment.

To provide some initial context, PVP, opposed to the TVP, demands end-user-centered product development. It thus focuses on a product’s value add in real life i.e. to what extent the product integrates into existing daily habits and workflows opposed to unrealistic controlled testing environments. The underlying and mandatory assumption is that users are lazy. You either nurture them and transform them into compliant human beings (which is rather challenging especially w.r.t self-induced illnesses) or you take into account sloppiness and lack of discipline when designing a DTx.

Observing a sluggish but certain movement towards value-based reimbursement, value generated in the real world (i.e. PVP) will evolve into the dominant KPI in DTx. Due to the central role of the PVP in the determination of success in DTx, let’s take a closer look at the determinants of a PVP:

Trust & Credibility 🤝

A. Theoretical value proposition 📘📕 📗

The backbone of any obtained credibility and trust of a DTx lies within an existing user-facing clinical value add, the theoretical value proposition. This can be transported to the users through:

• The tracking of indication specific healthcare outcome measures e.g. Hba1C for Diabetes or ISI for Insomnia
• User or health-care professional testimonials

Consequently, a (prospective) user might think: “In theory, if I adhered to the product specs, this would lead to an improvement in Hba1C of 1.5%.”

Remember, this functioning TVP only acts as a user acquisition instrument. However, a PVP is required to retain the users.

B. Personalisation 👩🏾 👴🏾 👵🏾 👶🏾 👱🏾

Personalised care is one of the central facilitators of care efficiency, enabling better health outcomes at (ideally) lower inputs, with regards to time and effort, often translating into lower monetary inputs. Any kind of personalisation is data-induced. During onboarding, the first data retrieval for the first level of personalisation occurs. Generally, there is an inherent trade-off between (a) higher initial degree of personalisation (and thus PVP) and (b) higher onboarding conversion rates. Especially in early stages, we strongly urge companies to focus on (a) at the cost of lower initial conversion rates. You cannot please every user group, so first focus on the low-hanging fruit, i.e. early adopters that are not intimidated by a more extensive onboarding process. The corresponding insights can then be used to create a more user-friendly and personalised experience for the more difficult and often broader user group.

Generally, we differentiate between two kinds of personalisation:

• Indication-specific personalisation: Every course of disease is individual and dependent on numerous external and internal factors. Often patients can be clustered into different patient groups e.g. into different skin types in dermatology. This ease of categorisation allows for a more straightforward and immediate therapy personalisation. Other illnesses however are hugely individual e.g. diabetes where my insulin need differs from that of my brother’s when we’re exposed to physical activity. This more granular personalisation requires greater amounts of user feedback and longitudinal datasets, to enable highly individualised and dynamic actionable insights and therapy plans. We call this delayed personalisation.
• User-specific personalisation: At the same time, users are not a homogenous group — their preferences vary significantly. This type of personalisation mainly revolves around UI/UX related personalisation. For example, users get to decide which colour, character or tone of voice they prefer. Ideally, this can be edited on a continuous basis.

Eventually, every DTx should aim for modular therapies with ideally multi-dimensional touch points, if required. While some patients prefer to use chat, others prefer to use video as a means of communication with health care professionals.

C. Emotional attachment 👦🏽💙❤️ 💛📱

It is crucial for users to take note of their indication and integrate it into their daily (indication-related) habits. We call this conscious inclusion. Obtaining this requires a certain emotional attachment between the user and the DTx solution. This can be obtained through:

• Gamification
• Benchmarking with peers
• Positive reinforcement

We have not yet come across any DTx solutions that neglected this category and succeeded in PVP and sustainable user engagement. No matter how good your TVP is, you need to bond with your users. Remember, users mostly are undisciplined and lazy. Converting them into power-users takes some “manipulation”. Check out Noom, who in my eyes have excelled in this dimension.

D. Privacy & security 🔒🔑 💾

In light of ever-tightening regulatory environments in today’s world, the below traits about privacy and security more act as underlying requirements rather than differentiation enablers. They include:

• Personal data protection through robust internal security management and monitoring systems (i.e. adherence to Europe’s GDPR regulations)
• Data provenance and source traceability: provide end users with a privacy notice that describes how the organisation collects, uses, and retains end user data, the types of data that the product obtains, the length of data retention, and how and by whom information is used → generally, the clearer this is communicated the better

Pragmatism & Intuitiveness 💡🤙🏼

A. Straightforward data entry 📝🗳

This is the centrepiece of any functioning DTx as it acts as the foundation for any therapy customisation and determination in the first place.

We often see DTx solutions that merely act as a passive one-directional content consumption tool. This does not suffice to sustainably alter human behaviour. Users need to provide their health companion with both initial and ongoing feedback for it to tailor its insights and advice.

However, oftentimes data entry is cumbersome. An example: Diabetes diaries require you to track all your blood sugar levels, physical activity, carbs consumption and insulin intake — all this being an absolute nightmare as it requires fundamental adjustments of daily habits.

Data entry needs to be pragmatic. Generally, the more passive (and automated) data entry/generation is, the better. This often occurs through integrations with hardware (as in diabetes or cardiology), leading to greater seamlessness. Alternatively, new categories of biomarkers can be used, including visual and vocal biomarkers such as the ones used by binah.ai and EVOCAL Health.

Ultimately, users entering data is conditioned to a certain credibility of the offering — an initial TVP often already helps. Users need to feel that it is worth the effort.

B. Straightforward data output 📊 📈 📉💡

Users need to understand the product to absorb and execute the proposed therapy. Remember that patients do not speak or understand medical language. Often, patients’ indication literacy is rather low i.e. they don’t know much about their indication. User overload is often followed by user surrender.

DTx solutions should not only focus on the end-user but also identify adjacent stakeholders that also form part of the patient journey and make sure their solution smoothly integrates with them.

One increasingly relevant group of stakeholders are health care professionals as they now represent a rather powerful sales channel, at least in Germany through the Digital Care Act (DVG). Hence why the doctors’ needs and preferences should not be ignored. More specifically, they require:

• No fundamental change in workflows
• No additional effort in prescription. Prescription must be as seamless, simple, and painless as prescribing an Rx. This whole process should not take any longer than two minutes.
• Once it is prescribed, the physician needs to have access to information on adherence, efficacy, and outcomes to improve the management of the patient.
• Overall, the use of DTx solutions should lead to greater revenue per working hour

Conclusion

Now it’s time to connect the dots. Below you can find the PVP flywheel:

• The foundation for any DTx is the adherence to privacy and security requirements as well as a TVP. Otherwise, fundamental trust building will be impossible. If the two are present, awareness in the respective target market is created.
• User onboarding needs to be characterised by a combination of emotional user attachment and initial personalisation (both indication- and user-specific)
• Straightforward data entry acts as the basis for recurring user feedback, creating the backbone of any delayed (data induced) personalisation.
• Straightforward, user-friendly data output coupled with delayed personalisation further increases credibility, acting as the basis for treatment compliance, leading to maintained data entry willingness. At this point, the PVP has been reached 🙌
• Generally, the more data has been generated, the better the product can be tailored to the needs of the user, further increasing trust & credibility. Once the flywheel gets going, it should be hard to halt.
• The logical consequence is ever increasing user stickiness and retention, leading to longitudinal datasets, being the backbone for the so-called monetisable state 🦾

In light of the underlying importance of the PVP in the determination of success in DTx, we are particularly interested in your secret ingredients of the PVP. Either reach out to me here on Medium or through jan-hendrik.buerk@btov.vc ✉️