Service quality evaluation through emotion recognition
Computer Vision technology for customer satisfaction evaluation
What was before?
Customer satisfaction is crucial thing in today's highly competitive market. A lot of factors contribute to the customers' happiness and loyalty. Service quality will be one of the most important of them. Is your staff mostly friendly and helpful? Or are they sluggish and indifferent or even rude and scare the customers off every day? Most of business around the globe hold most of their meetings online. Videoconferencing is used for meeting with customers and partners as well as for internal meetings, trainings and conferences. Many of them record their meetings for future reference. Companies have accummulated hundreds of meeting recordings and store terrabytes of videodata.
What was the issue?
Normally the managers overlook the service quality personally, but they can't do it 12 hours a day. Also, the managers are people too, so there may be human factor involved.
What did we do?
We have created an Artificial Intelligence boxed solution that recognizes emotions on the customer faces and measures their satisfaction and provides the analytics on the emotions that the customers experience standing at the cash desk and their rate of satisfaction, and also the number of customers and controversial situations.
What was the result?
With the help of the solution, the company has managed to define which members of the staff were providing service that induced more positive or more negative reactions from the customers and measure the service satisfaction and collect statistics across the entire chain. This allows to understand who of the staff members should improve the way they work with the customers. The company management can put more effort in training and even arrange the motivational program to award the best ones with bonuses. The solution can evolve and get additional gamification functionality in the future. It can be used to increase the conversion by recognizing the customers' emotions and offering them games, bonuses, discounts, shopping recommendations.
How it works?
Our Machine Learning models were running on the minimalistic operational system (Alpine Linux) with a web camera that is installed at the cash desk. As a customer approaches the cash desk and starts talking to staff member, the ML algorythms analyze the video stream, detect the customers' faces and emotions on them and send this data to the central server. The server collects the data from all cash desks and generates the analytics for the management. You can generate the report on the company level, region level and branch level. You can even drill down to the level of every individual employee. In order to do that the branch management has to register the working hours and the working places of the employees. This data will be integrated with the emotion detection results to show how the performance of every particular employee affects the customers' satisfaction.