Agirc-Arrco is a supplementary pension fund for the private sector, that supplements, by definition, the basic pensions paid in France by the Sécurité Sociale or the MSA. The Agirc and Arrco schemes merged on 1 st January 2019 into one scheme. Agirc-Arrco had over 22 million contributors in 2017 and decided to call upon dydu to implement a chatbot for their subscribers.
Maxime Garcia, Project Manager in the Digital Innovation Team at Agirc-Arrco, is describing the set up of the chatbot and its results.
What is your role within the organisation?
I am a project manager within the Digital Innovation team for Pension Product Management. I manage several projects focusing on customer relations and digital innovation, including the chatbot which is my main task.
What digital issues are you currently tackling, particularly in terms of conversational technology?
We aim to provide a range of homogenous services within the sphere of supplementary pensions,whilst improving the schemes’ brand image; responsive, innovative and readable.
Which problem(s) were you looking to solve by automating your user dialogues?
We receive numerous requests from our users, and they cannot be handled solely by our advisers. We needed to implement a chatbot to deal with the excess requests via automated responses. Our objective was to increase customer satisfaction while decreasing management costs, by allowing the user to be autonomous on the internet.
Why did you choose dydu?
Dydu covered all our specifications and requirements by offering a solution adapted to our needs.
What difficulties did you face and how did you overcome them?
We didn’t come across any particular difficulties. The few bugs we encountered were rapidly resolved.
What benefits have you observed since implementing the solution?
The chatbot has helped reduce the flow of calls to our customer relation centres. It has also made it possible to provide a customer service outside of the centres’ opening hours. The chatbot currently handles 30 000 dialogues per month on average, with an 85% success rate.
Do you consider the project to be a success? For what reasons?
The implementation of the project has been a success because our call flow diversion targets have been met, while maintaining a high level of understanding of our users’ questions.
How do you see conversational technologies evolving in years to come?
Automated conversations are becoming more and more essential to meet a company’s productivity targets. It is therefore necessary to rapidly develop these kind of tools.
How do you plan to remain innovative in this sector?
To avoid becoming obsolescent, we plan on continually improving our tools by closely following all the latest innovations in the field. We must seize all the opportunities that arise at the correct moment and stay tuned to the uses of tomorrow’s conversational technology (voicebots, callbots,