Businesses belonging to the hospitality sector include, among many administrative processes to start their business, the taking out of an insurance policy.

In many cases, this contracting takes place at the end of the whole process of entrepreneurship and means, for the person starting, yet another new procedure to complete that delays their objective a little more: opening their hotel business.

Streamlining and optimizing the process of contracting this type of insurance is key to being able to provide the quality of service that the customer expects. To do this, it was essential to intervene with a technological solution that would improve the user experience.

Total digitalization of the process

The hospitality industry is one of MAPFRE's most important business sectors, with more than 50,000 customers. For this reason, the company decided to carry out a field study to find out how the hospitality industry perceives insurance: what the process is like, how it is managed and what they expect from it.

The results showed that, for most of them, taking out hospitality insurance was a manual, tedious and complex process:

In addition, the outbreak of the COVID-19 pandemic made this process trickier due to the restriction of visits and mobility, making the process even more inefficient.

The main challenge was to encourage agents and clients to offload the more complex processes of management to technology by making use of digital underwriting and verification. All of this by offering specialized hospitality insurance, providing greater transparency to the customer in decision-making and improving the user experience exponentially.

Bringing agility through AI

The solution proposed by emergya was the creation of a platform that incorporated a series of Artificial Intelligence solutions characterized mainly by:

1. Agility, since it directed the process towards self-management by users, shortening times and optimizing the process.
2. Communication, that is favored by the omnichannel.
3. Transparency, as this solution simplified the process and helped to clarify technical issues for the user. 

For the development of this tool, several Google Cloud Platform AI solutions were used, such as AutoML and Vision, for the images classification and anonymization, and Google Maps. It has also been linked to other external public information services, such as the Electronic Headquarters of the "Catastro".

With this platform, a new 100% online risk assessment process is proposed. The use of Artificial Intelligence in the first contact with the insurance company not only guarantees agility but also generates value for the customer:

• The application can be used by both the customer and the agent whenever they decide.
• Geolocation and other web services allow MAPFRE to know the data of the environment and the business itself.
• The data standardization allows the value proposition to be adapted to the customer, making the most of the information.
• Image recognition allows complex procedures to be carried out by sending photographs and videos, thus facilitating risk identification.
• The automation of the process allows MAPFRE to provide an immediate and reliable response. 

In addition, all this is the starting point for the company to iterate the solution towards predictive models. In this way, they can facilitate the commercial work by offering agents not only more tools but also greater simplicity of use.

Rating time: 12 minutes

The tool has been highly valued by both customers and agents of the commercial network, being understood as "the insurance of the future". Only two presentation sessions were needed for users to quickly adapt to the new process due to its simplicity.

The most striking consequence of this new digital and automated process is that the risk assessment is reduced from one week to just 12 minutes. In addition, the need for sales representatives to travel is eliminated, resulting in a much more sustainable process.

Finally, with the incorporation of AI, the quality and organization of the data received has been improved, with a success rate of over 90% in the classification of images using machine learning models.