Advertisement

http://insurancethoughtleadership.com/wp-content/uploads/2014/04/bg-h1.png

Facebooktwitterredditpinterestlinkedinmail

October 2, 2018

Removing Language Barriers for Insurers

Summary:

When a broker is selling a life insurance policy--and can deftly communicate in the customer's language---sales go up.

Photo Courtesy of Pexels

One of the bigger issues at InsureTech Connect this year, I expect, is a result of advances in globalization and technology: How can providers more efficiently address multilingual needs without, say, a lot of Rosetta Stone? The challenges aren’t confined to insurers operating internationally; increasingly, geographic diversity presents this challenge within nations.

For example, what happens when an American customer traveling in Spain needs help urgently, but will be routed to the nearest call center–in Germany? And that office doesn’t have a translator?

Perhaps the matriarch of a small family business, run by non-English speakers in the Midwest, needs to purchase a policy? How can you upsell? Google Translate alone won’t work in such a complex situation; we need technology that quickly understands tone, slang and cultural context in addition to intent.

Even Facebook is just now getting into language-barrier solutions. How we communicate with potential customers affects the way they receive the information–and that can go well, or not.

This need extends to the point people for customer acquisitions: brokers. But after underwriting, it is the provider–not the broker–that does all the serving. If brokers aren’t adopting strong translation technology programs, those customers are lost. (Which is why we’re seeing insurance companies providing their brokers with materials in different languages.)

Luckily, advances in automation, machine learning and artificial intelligence have shown to cut costs, boost efficiency and improve capabilities at scale, across industries. For insurers, technology can now provide multi-language support through an automated customer service system, a capability we didn’t have just five years ago.

See also: 26 Most Important Words in Business  

In legacy automated customer service systems, the program looks for individual keywords within an interaction. But today, technology can absorb 100 paragraphs–a library of text that reflects a displeased customer. Agents can even, for example, see whether the customer hung up mid-call, or used profanity.

Simply put, each of those 100 paragraphs is assigned a numerical number–some as long as 100 digits, each with a label that reflects those human details that can get lost in automated service–anger, happiness or anxiety. It’s a sentient technology.

Codifying language using a binary library speeds up processing time by making it easier to find complex combinations of words, as opposed to the previous method of searching for keywords.

Another weakness with the traditional keyword search is that it can throw off algorithms. If a customer declares, “I’m red in the face with your service!” the program will not be able to interpret it for what it is–especially when translating across languages.

Insurers can no longer keep up with the old model of staffing a few well-versed interpreters now that people opt for digital platforms, like text. That’s largely because consumers are opting for text-based communications, like email or messaging apps, over call centers. Translation needs have shifted to digital text, which demands an entirely new solution.

Artificial Intelligence can digest a call in one language, translate it and transcribe it as a ticket–even assigning a score to each conversation. This score is important, because it determines the routing and triage process. If Company X decided to challenge itself and escalate any conversation that scored under 80%, the company can do that–and it can all be automated.

For the past few years, Microsoft has been on the forefront of this sector, but still has drawbacks: Users must train the system over time, because the program is industry-agnostic in its application. When this same cutting-edge technology is built exclusively for the insurance industry, it becomes far more powerful. Over the last three years, we’ve seen that 90% of customer questions will be the same, with only 10% unique to the system’s “brain.” That promises better accuracy and a smoother move across the pipeline.

See also: Language and Mental Health

Because this is a translation issue, it’s no surprise that we are seeing a lot more demand for multilingual support in the travel industry, particularly with EU carriers, which serve a diverse range of demographics. A close second is demand we’ve seen in small business property and casualty insurers. Research has show that Russian is an important language to accommodate, as the population is traveling more.

The value of investing in advanced translation technology is transparent. Policies are sold in a process that involves around eight touchpoints–the referral from a friend, the initial phone call, the in-person visit, the medical exam, follow-up calls, to name some. Customer experience determines wins and retention. When a broker is selling a life insurance policy to someone–and can deftly communicate in their language—it makes sense that customer affinity (and thus their sales) would go up. Add in the ability to do all of this across platforms–messaging apps, email, chatbots, Alexa or social media–and it will make a difference in the bottom line.

description_here

About the Author

Aamer Ghaffar is CSO of Kordinator. He has almost 20 years experience in technology business leadership, primarily in healthcare and insurance.

follow @ITLupdates for more stories like this

Email Aamer

+ READ MORE about this author ...

Like this Post? Share it!

Add a Comment or Ask a Question

blog comments powered by Disqus
Do NOT follow this link or you will be banned from the site!