Having a grasp on customers’ emotions is never easy. Nowadays, traditional subjective methods for this have been giving way for Artificial Intelligence, which allow businesses to improve their service quality and win the heart of customers.

Businesses can only prevail if it wins customers, so understanding what customers think and bringing them best experiences is always a top mission for all business leaders. To grow, it is best that businesses focus on satisfying their customers. Seeing that the development of technology and science has been closing the quality gap between corporations, it is customer experience that plays the key role in determining business competitive capabilities and maintaining customer loyalty.

In 1986, Parasuraman and his partners had created the SERVQUAL model for service quality evaluation to help businesses eliminate the expectation gap between customer expectations and real experiences.

This model is widely used in various nations, and of course isn’t unfamiliar to service quality managers, but one disadvantage to it is the complexity in evaluation. Many studies had demonstrated that businesses that excel in satisfying their customer all share one common point – they have specific evaluation models, detailed staff training procedures, and frequent model reviews via mystery shopping.

Will the fruits of the 4.0 industrial evolution, or AI developments in particular, actually assist business leaders in wining their customers’ hearts? The answer is yes, and AI will even directly address the many questions in the Parasuraman model in specifics with real figures and little subjectivity.

Knowing your customers – Who they are and what they want

Before, customer data only involves shopping history and some general personal information of questionable accuracy, stored on business CRM systems. Today, however, social networks have allowed businesses to know much more about their customers.

In 2015, our team had built a Social Listening solution for a retail business based on Twitter data along with IBM Watson algorithms, which can generate in-depth analytics regarding customer insights alongside basic information.

Tweet analytics for knowing customer personalities.

With the solution, whenever a customer steps into the store or publish a comment regarding a product, employees will receive information regarding their preferences and personalities to decide on a suitable and personalized approach, thus bringing about better services. The end goal, however, is still to persuade customers to buy, but not making them feel confronted with cliché contents.

Better service quality management

Is it possible for management to ensure compliance with service standards in employees miles away? Will this compliance become a habit, a pride, or just a front to deal with campaigns and checks? The mystery shopping method is often costly and time-consuming, but AI will provide more options for businesses.

Today’s camera systems have got smarter thanks to Deep Learning and its major breakthroughs in computer vision.

Store traffic and sales are always vital in evaluating employee capabilities. In the past, store managers might be too far away, and the only explanation they might have received for the lack of sales was that there was no customers. Meanwhile, solutions for store traffic tracking are very common and accessible nowadays – thanks to AI and the YOLO model using CNN (Convolutional Neural Network) structure, both with high accuracy.

An AI application that can count the number of customers inside a store.
Supervision inside a restaurant.

Furthermore, the rise of SNS alongside online shopping has made it easier for us to access opinions and evaluations in forms of Facebook and Twitter statuses and comments. However, are businesses really listening on these platforms, or are they too dependent on traditional channels like call centers and market survey campaigns?

There is also the edge-cutting NLP technology in these recent years, which has allowed for more accurate sentiment analysis results, helping businesses in evaluating campaign effectiveness and determining preparations for possible negative events.

Sentiment analysis results based on SNS keywords.

It is also of great necessity to build experience maps for customers as they approach a service, especially as businesses often disregard post-sales – a vital step in such maps. How to evaluate if call agents maintain pleasant attitudes to customers, even in front of anger or annoyance? For this problem, we often record conversations and listen to them randomly. However,  AI applications have been able to detect call abnormalities, allowing better management and timely adjustments to improve quality.

Sentiment analysis results based on SNS keywords.

Above are projects that I have personally been through. Of course, that’s not mentioning all the tales behind those famous for using AI to understand customers like Facebook, Amazon… I also believe that the time is right for businesses to quickly approach AI and enter the race to understanding customers.

Khoa Tran – Solution & Technology Unit, FPT Software

Related posts: