Why healthcare is starting to look at conversational interfaces

I recently attended a chatbot meet-up put on by Delta Dental of Washington, featuring Lili Cheng, Corporate Vice President of the Microsoft AI and Research division, who spoke about AI and conversational interfaces. Conversational interfaces, such as chatbots and voice assistants, are a hot topic right now, and we’re seeing many of our clients trying to wrap their head around these technologies and how they can use them to drive down their costs while improving the customer experience.

This technology is rapidly maturing, and the business use cases are still being tested. We’ll discuss some of the challenges and shortcomings in this blog. However, overall, we are seeing lots of excitement and initial success with conversational interfaces.

The Way of the Future

Over the past two decades, we’ve seen amazing leaps in technology, especially in how humans use computers to accomplish tasks and find information. Cheng posed during her presentation that customers will continue to expect more and more out of technology over time. User interfaces are not absent from this demand by any stretch of the imagination. We’ve seen the evolution from “ancient” green screen systems to touch UI; now we are headed into the age of conversations as the primary UI, and organizations need to be ready to deliver this experience.

How So I Use It?

The first place that I’ve seen companies focus on when looking to implementing this technology is all around customer call avoidance. In theory, this is an ideal technology to solve this issue and drive more self-service, but the truth is the technology just isn’t quite there yet to deliver nuanced explanations to highly complex customer questions. We use Siri to check the weather, Alexa to add something to our cart, and KLM’s Facebook chatbot to look up flights, but these are all straightforward use cases. Conversational interfaces generally work best for simple transactions like these, with simple results and workflows that are structured, narrow, and relatively predictable. This makes overall call avoidance more difficult, but there is a lot of room for call efficiency.

Let’s look at two examples in health insurance. For call avoidance, providers are a great fit for a conversational interface. They speak the insurance language and have quick transactional questions that are ideal for conversational interfaces, such as determining member coverage by procedure code. By giving providers access to this new interface, you could drive some significant volume reduction from this “customer” group who are primarily seeking instant answers to relatively simple queries.

In terms of call efficiency, members are likely an excellent fit. By using the bot to capture member info up front, your representatives can focus on higher value activities prior to addressing a member claim question. The conversational interface would authenticate the member and capture their claim details, which are then presented to the representative prior to engaging with the member – validating and retrieving the appropriate information in advance allows for a seamless customer experience - preventing the member from repeating their question, and saving the agent time in finding the information. Though these technologies usually allow for some level of escalation to a customer service agent, in time we expect conversational systems will be able to handle much more complex use cases.

What’s the Tech Look Like?

When it comes to conversational interfaces, there are many different pieces of the technology puzzle that you need to bring these experiences to life. Adding to this complexity and noise in the market, there are lots of different players coming out of the woodwork trying to provide all of the different pieces to this puzzle. Microsoft may be close to delivering a fully packaged solution, but there hasn’t yet been one dominant market leader delivering a true end-to-end, plug-and-play platform. Looking at the technology stack, the pieces that are unique to conversational interfaces are the channels (e.g. Alexa, Facebook, others), the natural language processing (NLP) tools (allowing the systems to understand a user’s intent), the underlying business logic and workflows, and of course the systems integrations required to build these interactions and tie everything together. As you approach this technology, evaluate each piece of the conversational system puzzle and keep in mind how each component fits into your overall technology stack as you would any other digital channel.

Where Do We Start?

We know customer expectations are continuing to shift: disruptors such as Amazon and Uber have set the baseline for customer experience, and customers expect a seamless, multichannel experience from their health care providers and insurers. AI and chatbots are part of the mandate for healthcare companies to embrace a digital future, and you need to act now to build out your own foundational capability. This requires a couple things:

• Start with the data - The usability and accessibility of data that will feed into your conversational assistant is by far most important element of your ability to get started. If this data isn’t exposed on your member portal or mobile app today, it’s going to take work to get it ready for a bot. This channel is no different to your other digital channels you’re using to engage customers today.
• Identify your use cases - Keep in mind for whom, what and where you are delivering a bot. What customer segment are you trying to engage? What are the main issues that you can address with a conversational UI? What are the primary channels you will use to engage these customers? Each of these three questions needs to be considered in reference to each other.
• Build, test, and learn - You need to develop a proof of concept and present it to a user group, may that be internal or a beta customer group, and be open to what results you find. There are some best practices out there, but until you get your hands dirty and publish a bot you don’t really know what you don’t know. Internalize those learnings and evolve your bot to deliver value to your customers.

If you’re looking at conversational systems and chatbots to solve your problems on day one, you’re getting ahead of yourself. To stand this up this will require you to look at conversational UI as its own business capability and will require investment to effectively support it. Having the willingness to fail is key--then take the time to reflect, learn and shift your focus going forward. Act now to build your capability so you can evolve with the technology and customer demands over time.

If you have questions or would like to talk about how West Monroe can help you plan, design, and build your Conversational user experiences, please contact us.