Saturday, April 16, 2016

Data Science Can Empower Benefits Leaders Communicate Less and Yet Accomplish More

In many companies today, internal communications overload is a big problem. Internal communication teams including human resources teams are wary of sending out more emails or direct mail than necessary. They also know that most employees ignore internal communications because such communications are poorly written, poorly designed and not personalized. For critical communications such as benefits enrollment, many human resource departments use electronic mail, paper mail, and posters to ensure that the message gets across. Most employees have learned to ignore internal communications unless that message is repeated multiple times.

In this scenario, if a benefits leader wants to educate specific segments of people about health benefits that are relevant to them, there were limited tools available to them. My team at Castlight has taken significant steps to address this need among other things. My purpose here is not to promote the products we build. Instead, I want to talk a bit about how we are planning to use the lessons learned while building Castlight Action, our predictive analytics and personalized recommendations product,  to empower benefits leaders in all our customer organizations, make employees aware of their benefits, to enable employees become healthier and even help them save some money in the process.

Since we have data about the behavior of millions of users and billions of claims, we use data mining techniques such as cluster analysis to identify segments of people with similar attributes who behave the same way. We may not understand why certain segments of people with similar attributes behave the same way, but we know that they do. Then we identify the people with such attributes who are not behaving the same way and tell them they might want to consider such behavior.

The above paragraph is very abstract. So, let me share an example. From our data, we know that for some reason people in their twenties care a lot about dental health. In other words, people in their twenties get dental checkups every year. Once they reach their thirties their lives probably get busier with family and children and they do not get as many dental checkups. So, let us say they were a 1000 people in their twenties in a company with 10,000 people. 800 out of those 1000 people get dental checkups every year. So we wonder about why those 200 people in their twenties are not behaving like the others. Maybe they just joined the company. Maybe they don't know any good dentists near their home. Maybe they think it would be very expensive to go get a dental check up. We really  don't know why. However, we know that if we tell them about the dental benefits they have via their employer and enable them to find high quality, reasonably priced dentists near them, there is a very good chance that they will get their dental checkup done. This leads to benefits awareness, benefits satisfaction, better health, better productivity and in many cases some cost savings.

The interesting thing is that we can accomplish this by communicating to just those 200 people rather than communicating to the 800 people who already get their dental checkups and the 9000 people who are not in their twenties and don't care about dental health. I took a fairly simple example to explain how we reduce the need for mass communication about benefits and significantly increase personalized communications that engage employees and slowly but steadily change behavior, improve benefits satisfaction, make employees healthier, and even save some money for the employer and the employee. Most segments are not this simple to build or comprehend.

Building such an infrastructure and assembling a product, data science and communications team that can do this can cost employers tens of millions of dollars, even if they have a clear idea about how to do this.

Today, Castlight has several such segments and serve many employers in the US. Every month, we keep adding more segments that make personalized highly relevant and valuable recommendations to employees. Every piece of communication we send out makes one person a bit more knowledgeable about his or her benefits, has the chance to make him or her a bit healthier, makes him or her a bit more productive and saves him or her and their employer a little bit of money. We do this several million times every year. Knowing this is what motivates me to get up every day morning and go to work. If you are a physician, data scientist, product designer, content strategist, or a product manager and think this is fun to do, give me call. I am on LinkedIn.

Related Posts Plugin for WordPress, Blogger...