Case Study

How a language neutrality engine helped overcome language barriers for a European insurer

Print This Page

How a major insurance company successfully centralized and automated key claims reporting and policy administration processes in multiple languages using Genpact’s language neutralization engine.


Business challenge

Is it possible for a company to optimize customer service in multiple languages? That was the challenge faced by Tryg which is one of the largest insurance groups in the Nordic region. The company sought out Genpact to help resolve this dilemma.

Tryg wanted to launch an online channel to handle claims processing and policy administration for its customers in Denmark, Norway, and Sweden. The company’s executives were surprised to learn that Genpact’s language neutrality engine could enable the centralization and automation of many business processes, improving the customer experience while cutting costs.

The need to administer policies and to process claims in multiple languages was a major concern because it limited the productivity of claims teams, increased response times, complicated customer communications, and led to a proliferation of high-cost, non-standard procedures. The company’s First Notice of Loss (FNOL) process was the biggest hurdle, since it relied on voice-based customer interactions. Tryg wanted to improve service by moving customers to a digital channel that incorporated web, email, and live-chat interfaces. But delivering this service through a centralized team meant integrating processes for classifying, inducting, and assigning daily communications across 14 departments in three countries.

Figure 1: Process flow in FNOL registration

SeeClose detailed view

Genpact’s solution combined state-of-the-art technology, business process expertise, and the domain knowledge of an experienced workforce

An automated translation tool was an essential component, but it required a staff that was specifically trained to use it in order to fully interpret industry-specific expressions in Swedish, Danish, and Norwegian. It was also important to quickly identify exceptional cases so that Tryg’s in-market customer service employees could focus their efforts on situations where personal interaction was the most critical.

Genpact implemented a language neutrality engine in three layers. First, glossaries for routine and templatized operations enabled a centralized team to quickly learn to efficiently handle the bulk of customer interactions for policy administration and claims. The second layer incorporated Systran’s translation tool. And the third layer — “smart scoping” —identified which processes could leverage the translation solution and be serviced centrally. To implement this third layer, Genpact documented workflows and analyzed procedures to determine the degree to which each one was language dependent. Language experts were brought in to interpret the most complex cases.

Take a copy for yourself

Download PDF

At first, Tryg executives were skeptical that automated translation could deal efficiently with the wide range of idioms and specialized terminology used in its business.

“We did not fully grasp the potential up front, but we’re realizing it as we go through the journey. The high proportion of processes that can actually be covered with such simple tools was surprising to us.” 

Continuous improvement through machine learning is the key to this success. Machine learning enhances the underlying rules-based translation tool so it becomes smarter each time it encounters repeated words and phrases. To expedite the learning process, for several months Genpact experts analyzed transaction data, engaged language specialists to refine interpretations, and updated the engine to the point where it became self-sustaining. 

For Tryg, delivering a great customer experience through the digital channel was the most important outcome

Optimized claims handling and policy administration streamlined communications and enhanced customer satisfaction. Technical support was improved and process standardization led to increased coverage. In addition, efficiency improvements and the use of a centralized team to service a substantial proportion of transactions generated large cost savings.

The story doesn’t end there because of the continuous improvement inherent in the language neutrality engine and the team’s ability to continue mastering it for ever-more-complex interactions. Mr. Klimfors says that Tryg recognized from the start the need to balance two objectives: improve the customer experience, and realize efficiency gains from centralizing common processes. “Those were the criteria we used,” he says, “and we’ve seen over the last year and a half that the needle is moving on those criteria all the time. And as our comfort level increases, we’ve seen more and more operations being serviced from an offshore location as language neutralization gradually increases in power.”

Continue Reading

Ready to
Explore?