Fitbit loses 5% of users within a week of purchase and 12.5% within a month. Fitbit hasn’t overcome the human tendency to quit lifestyle changes despite having a revolutionary product that yields long-term health improvements.
But what if I told you that Fitbit has, comparably, an amazing retention rate?
Patient retention is a problem for much of healthcare
Virtual care systems, such as telemedicine, have a patient retention challenge as well. Remote care – much like remote work – requires motivation effective enough to overcome our tendency to quit.
The tendency to quit is strong and goes across our health care experience. For example, the percent of people still taking their statin medications 90 days after getting a prescription is only around 60%.
This challenge of people quitting is amplified within B2B health tools, like Conversa, because we must keep multiple parties engaged. Doctors, administrators, and patients are just some of the stakeholders B2B tech must delight.
Digital health technology depends on patients’ actions
From opening a daily meditation app to logging into different healthcare sites, the technology is dependent on the patient’s action (or inaction). When the patient is feeling good, there’s more of a drop-off.
In my observations of chatbots, retention is really driven by two things:
- Who the chat is coming from (do you trust this person?).
- Is the topic of interest to the patient?
If a patient answers yes to both of those, retention will be higher.
Related Content: How Chatbots are Helping in Healthcare
Improving patient engagement
Here are a few ways we observe virtual care companies, including Conversa, working to improve patient engagement:
1. They use language strategically
Not only do they use language strategically but they ask questions that are easy to answer.
If you think building a chatbot for health care is about technology, think again. At the root of these interactions is language. There’s a conversation that is going to decide what happens next for the patient.
In order to have a wider view of how we hold dialogues, we hired writers with unique backgrounds. Our director of content strategy is a Nickelodeon alum who ensures the language we use is approachable and empathetic.
We work with psychologists to make sure that our words support patients’ self-determination.
2. They focus on the quality and effectiveness of content
Some companies are starting to shift their focus to the quality and effectiveness of the content.
-It aligns with patient autonomy
One way to approach this is to align what the bot says with a sense of patient autonomy. The bot speaks to the user in a way that acknowledges their competency in managing their care.
-It is relatable
Relatability is another important ingredient. It’s not surprising that patients open up to their doctors more if they feel understood. These are all tenets of Self Determination Theory, or SDT, which we embrace fully at Conversa.
-It is succinct
The other hurdle is that bots need to compete within an increasingly bombarded attention span. They need to make sure that every interaction is helpful. It also must be brief – 150 words or less.
To solve that, we ask for feedback after every interaction, and we are short and sweet about it: Was this interaction helpful?
As we refine the bots based on this real-time data, we’ve seen results. 97% of Conversa chat sessions are rated as helpful. This is true across regions and uses cases as different as lifestyle coaching, post-discharge after a hospital, before or after surgery, or in the management of a chronic condition.
3. They make access easier
The more advanced virtual care technology becomes, the easier it is to lower barriers to adoption and retention. It lets us “meet people where they are.”
We find it essential to engage the caregiver when the patient is “digitally unable.” That effort helps push our Medicare participation rates to over 50%.
This is crucial to the “am I getting value?” aspect of lasting tech. It puts caregivers in direct contact with their loved one’s health.
Another tactic is to offer easy access to a vetted library of interactions. This also goes hand-in-hand with self-determination theory.
Patients exploring their condition on Google may find some reliable but also blatantly false information. We provide the patient with the correct information directly.
4. They build in next-level personalization
AI and data analytics have made more refined personalization possible at scale. However, it is also important to personalize the right way. (Hint: starting an email with the person’s first name is not enough!).
This can be tough. There are a ton of data points in a patient profile, including:
- where they are in terms of their condition(s)
- their diagnosis
To keep patients on track with the platform, the bot needs to adjust each dialogue, every step of the way. And, it needs to be done at exactly the right time.
Conversa utilizes a variety of techniques to accomplish this:
-A taxonomy-driven profiling system
We use a taxonomy-driven profiling system to look in real time at patterns of clinical and non-clinical patient profile characteristics. This allows us to deliver exactly the right chat experience.
Every patient interacting with the Conversa chatbot gets a different dialogue. And, they never have the same chat twice.
Not only is this tailored to their specific issues, but it keeps the conversation fresh. Bots can do this particularly well.
When the bots will initiate a chat, is a very influential aspect of sustained engagement. It requires personalized timing.
Take Facebook, for example. Desperate to get their users back, people criticized them for innocuous push alerts, further repelling them from the platform.
The frequency of the chats needs to be calibrated to the patient’s own sense of urgency. If a patient has just been discharged for heart failure, we are going to check in on them more often. If they don’t respond, we remind them.
We have seen a 60% retention rate in our lifestyle cases–those who are losing weight or managing diabetes–after nine months of utilizing the chats in employer settings. We feel a good portion of that success has to do with our personalization approach.
Another strategy we use is to anticipate everything that the person we’re writing for may encounter at different points in their journey and in different circumstances.
For example, we know what someone who just undergone hip replacement surgery will have to deal with. So, we offer specific pre-surgery advice: buy a grabber, since you won’t be able to reach for things from the bed.
For head and neck cancer patients preparing for radiation therapy, we advise them that they will need their fillings removed.
We let them know that they may lose their sense of taste, temporarily or permanently. We also educate them ahead of time that swelling in one leg is not something to ignore.
This timely, personalized, anticipatory advice is what keeps them engaged. And, it reduces complications.
5. They understand the importance of patient trust
Patient trust is becoming a bigger area of focus when it comes to enhancing retention. Users need to be able to trust the chatbot the way they trust their doctors.
We see the platform as an extension of their (human) care team. We’ve seen patients express thanks for the bot interaction the same way they would thank their doctor. And it’s certainly not because the patient is confusing the two.
6. They integrate with other health tools
Integration with other health tools is important. Virtual care companies realize that if they can easily weave their functionalities with tools like Fitbits and Apple watches, the chances of keeping that patient engaged are higher.
Integration with the provider’s workflow is even more challenging and even more important. By integrating these technologies more seamlessly with the systems of care, they can be a more effective extension of that care.
The bottom line
To be successful in achieving meaningful patient and provider retention, virtual care and chatbot companies cannot just throw new technology at the same old problems.
They have to understand the features that really support patient retention. And, they have to focus on them with precision.