Introduction

Machine learning models enabled by artificial intelligence (later referred as AI) complement the work of designers by efficiently analyzing vast amounts of quantitative data. Based on data collected from customers, AI is able to create various alternative hypothesis on what kind of impact design can achieve. By modeling customer behavior machine learning models can predict which combinations of various possible qualities have the greatest potential for improving customer experience.

The evidence of Design 2.0 describes a collaboration between a European service design agency and a Nordic life insurance company. Machine learning models predicted that feeling of caring will have the most significant impact on customer satisfaction, customer loyalty and NPS (Net Promoter Score) of the life insurance company.

Client scoring data was totally anonymous which meant neither the client or  AI could identify a customer.

After AI had forecasted that the focus of the design should be in implementing feeling of caring, a group of service and business designers could start work towards putting this into practice.

They utilized an iterative and agile double-diamond process for concretizing the feeling into service actions and co-designed a new service identity together with management and employees. The service identity took into consideration the multichannel environment (digital - physical - social) in which the life insurance company operates and serves its clients. Finally, the new service identity was translated into an effective tool and guidebook that put the “feeling of caring” into
practice across the whole organization.

Process

At the heart of the new Design 2.0 lie innovative machine learning models, which build the basis for service design and ensure the impact of design decisions. Thus the process was started with a deep dive into the customer experience data available at the organization. In detail, the existing data analysed included customer satisfaction scores, employee satisfaction data, data from customer encounters and net promoter scores (NPS) from the past years. Furthermore, the
analysis included outside data that might impact customer experience, such as the stock market development (Nasdaq) from the past years.

It took approximately two weeks to gather all the quantitative data from various data sources and to complete a sanity check to ensure that the data is applicable for AI analysis. After this, the AI started to go through the data and learn how data points interacted with each other. The main task for the AI was to identify what is the correlation and possible conversions between the gathered data, customer loyalty and the net promoter score.

One of the first things AI learned was that, contradicting the general belief within the organization, changes in the stock market had no relevant impact on their overall customer experience. There was no correlation identified between the development of stock prices and customer satisfaction scores. The analysis proved, that this was just one of the cultural myths developed by the organization about their customers that could now be disproved and dismantled.

The AI pointed out was that there were three significant experiences that had the most significant correlation with customer satisfaction: the perceived expertise, the feeling of caring and the fluency of the service. Of these, the AI identified that the feeling of caring has twice as much impact on a positive customer satisfaction than the perceived expertise. This was another cultural myth proved wrong. AI also identified that the ambition of the service promise also has a correlation with the development of the net promoter score. AI analysis validated that if the service promise is set too high or is over-ambitious, it is nearly impossible to reach the level of excellent customer satisfaction.

Based on the results from the AI analysis, the service design team commenced qualitative interviews to understand the themes more in detail and to prepare them for translation towards a service identity. After this, a series or employee cocreation workshops were held daily during one week at the premises of the life insurance company. These “pop up” co-creation workshops were facilitated in a manner where it was easy for all the employees to join and give their input to the development.

Through the co-creation workshops, the service identity started to take its form. The Feeling of caring as a “soft” design driver became an extremely interesting contrast to the existing bold and masculine brand of the life insurance company. The brand of the company had so far been considered as intriguing, affluent and dominant. The service design agency recognized that these contrasts actually create a unique mix when a soft and empathic feeling of caring meets the masculine and powerful brand image.

After the co-creation workshops, the agency began verbalizing, visualizing and finalizing the new service identity. The service identity was given a name and a character. The employees were given concrete examples and scenarios how the service identity can be visible in the multichannel service environment and how the service signature can be adapted to the digital, social and physical channels.

The service identity was concretised as a guidebook and an operating model for introducing the feeling of caring in all customer encounters and service channels. The model presented different customer profiles and how the feeling of caring should be flexibly modified based on the customers’ specific expectations. The guide book also defined how the organization can as a whole support employees with a human-centric culture that nurtures the feeling of caring.

Outputs

After the service identity took a visual and written form, the life insurance company started piloting the model and emphasizing the feeling of caring with its customers. The first department to pilot the new service identity was the customer service. Supported by the design agency, the customer service team started by re-thinking some of their key processes and touchpoints: the service recovery process and the service process which starts after the death of a customer. Of the new ideas generated were more caring ways of apologizing after a mistake. These could include a personal phone call, a letter or a flower delivery – the key was that the apology exemplified the feeling of caring and the new service identity.

The process related to the possible death of a customer had always been difficult for the customer service personnel since the situation is so sensitive. At the same time it requires specific actions from the close relatives, such as closing or adjusting the current agreements with the life insurance company. In the pilot phase, the customer experience team created new ways for how to take a customer’s emotions into consideration. The emphasis on feeling of caring gave a permission for the personnel to take in more consideration the the feelings of the customers – and as well their own feelings. This made the handling of the sensitive matter easier and more natural for the personnel. The feedback received from the customers was instantly warm and very encouraging as the customers immediately sensed that they are truly cared for and the that life
insurance company understands their situation.

After a successful pilot in the customer service unit, the service identity implementation was expanded towards other areas in the organization. To ensure an inspiring and empowering way of learning the team utilized gamification. A series of workshops was arranged in all units and locations in which the participants played an specifically designed service identity board game. The purpose was to encourage the employees to take the ownership of the new service identity and to discover what the feeling of caring could mean for their customers and for themselves in their daily practice.

Besides the gamified implementation, the service design agency also created guidelines for continuously managing the new service identity. In a co-creative manner together with the managers of the life insurance company, the agency designed the “Year Clock”, an annual calendar, to communicate the management of the service identity during the different phases of the year.

The year clock was designed so that the new service identity activities are coordinated with other actions and phases taking place during the year in the organization. The guidelines for the management and the year clock also included updated criteria for the recruitment of new employees. The feeling of caring became something that the organization now wanted to be emphasized during the recruitment process as a quality for desired candidates.

Another critical touchpoint of the new service identity were the multiple documents the life insurance company sends its clients. The current documents did not exemplify the new identity, but instead were very traditional, and in some points, they even reduced the feeling of caring thus reducing customer satisfaction. The various documents sent to the customers were reviewed and a new, aligned document concept was co-created together with the employees and customers.

Finally, together with the law department the designers reviewed the changes contents for validity. This collaboration resulted in the law department getting also excited about the design process of the documents and they became very positive towards the approach. The outcome of this legal design process was a document design guidebook, which included a visual concept describing the most user-centric documents on the life insurance industry. The guidebook also
included concrete examples how their renewed documents emphasized the feeling of caring and manifested the new service identity and the brand of the company.

Impact

The sum impact of the design actions coordinated around the feeling of caring can now be clearly witnessed in the responses from their customers and their net promoter score (NSP) has steadily risen since the new service identity was launched. The life insurance company reached the exact effects they aimed for and they can now prove it with straightforward data from the NPS.

However important the NPS data is for decision making, the most essential impact for the life insurance company is that their customers are now encountering a renewed service that makes them feel cared for. Based on collected face-to-face and written customer feedback it is clear that customers have warmly welcomed the new signature in the service. Besides of the significant growth in customer satisfaction, the business year of 2018 was excellent for the life insurance company – the amount of wealth their customers trusted in the hands of the asset managers of the company grew record high.

The service identity project was able to create a common language and a joint signature service that is now visible and tangible across the organization. The collaboration between the European service design agency and the life insurance company started during 2017 still continues to this day. The agency has become a trusted partner in all things related to their customer experience development and to the larger cultural transformation towards increased customer-centricity.

Conclusion

The evidence of Design 2.0 shows how machine learning models and AI can compliment the design process. By being able to analyze immense amounts of data and finding correlations that are practically impossible for humans to identify, AI guides design resources towards the right direction. This ensures that the capabilities and skills of designers can be used exactly where they are needed the most. Machine learning models can also predict the business impact of design
more precisely and significantly support the value growth of design. When the business impact of design can be forecasted beforehand and measured afterwards, the value of design is no longer abstract or vague.

By combining machine learning models enabled by artificial intelligence with high quality service design work we can significantly build credibility towards design as activity with a high business impact. This project identified that the key qualities that AI can predict are customer loyalty, customer churn, customer satisfaction, employee satisfaction, tendency to use a service and advocacy (NPS). These form also the key targets in most of service design projects aiming to improve the service business for companies around the world. At the same time, the common trust built with the analysis can give designers more space for creating excellent services around the agreed theme.