Why TPAs need to embrace virtual assistants
A chat bot is perfectly equipped for a range of low-level customer service functions, such as confirming and gathering information.
It’s easy to understand this trend’s cause: the average cost of a call center employee is more than $1 a minute, and if that employee is trained in health care, the costs are more than double that rate. Our chatbot handles over 80 percent of member interactions with a 98 percent satisfaction rate, so you can see why this technology is so powerful. The typical customer service department spends time on routine things that a bot is perfectly equipped for, like confirming and gathering information, rather than performing higher level tasks
Related: Dave Chase on how AI will change the employee benefits industry
The other thing that virtual assistants are great at is outbound outreach campaigns. Right now, TPAs might perform limited outreach campaigns to the top 1-2 percent of claimants, but what about the other 99 percent? Health care happens to everyone, so how are you proactively getting in front of tomorrow’s claims rather than reactively dealing with yesterday’s? With the combination of claims analysis, which uses AI to identify trends, and virtual assistants to perform outbound campaigns, TPAs can address their entire population and do it often. Being able to reduce customer service costs while at the same time lowering insured spend is a winning combination.
In this same report, Gartner also forecast that 20 percent of brands will abandon their mobile apps by 2019. Many companies are finding that creating a mobile application is tough and getting customers to engage and adopt an application is not worth the struggle. Most companies are reporting that their in-house applications are “missing the mark” due to the cost of maintenance, support, upgrades and marketing the app. Many of these brands are investigating using chatbots within apps like Facebook Messenger or Slack, but due to HIPAA rules, this is out of the question for TPAs. You also need to make sure the platform you rely on is aligned with your business model. Just remember when Facebook reduced engagement to fan pages by 98 percent and started charging to speak to your fans.
Developing your own virtual assistant
Depending on the scope of the project, developing a virtual assistant can take a few weeks for a single developer or can take years. I would estimate that we’ve spent over 120,000 development hours creating our virtual assistant platform, but a simple project could be developed in a few months, and you can even have something to test in a few days. Before you start developing, you should identify three things:
- What are the opportunities to use a virtual assistant – Maybe your customer service spends 5 percent of their time dealing with supplying different vendor information and a chatbot could handle this task instead.
- What goals are you trying to accomplish – Defining key performance indicators (KPI’s) for your goals is essential.
- How do you want to build your chatbot – You can use custom development, programming frameworks, or development platforms (no-coding).
Custom development is the best approach for virtual assistant development, but it’s the most complex. This is the most flexible approach but not for the faint of heart or small teams.
For those that need more flexibility but don’t want to commit to a custom solution, there are frameworks by Microsoft, Wit.ai, and API.ai and a few others. These frameworks have a set of predefined functions and classes that a developer can use to speed development and not have to create everything from scratch. Development platforms like Chatfuel, Chattypeople and Motion.AI can be used to create simple virtual assistants with no programming. Your capability is limited, but you can test something reasonably quickly without a significant commitment. Lastly, for TPAs that don’t want to commit to any development or maintenance, partnership with existing healthcare virtual assistant providers is a turnkey approach.
Rick Ramos is the chief marketing officer for HealthJoy.com, an employee health care guidance platform that uses artificial intelligence and virtual assistants to lower health care costs.