Case Study: Employing AI to Personalize Customer Experiences

PinIt

Swisscom is building capacity to offer customized, real-time customer experiences through artificial intelligence.

Name of Organization:  Swisscom

Industry: Telecommunications

Location:  Ittigen, Switzerland

Opportunity or Challenge Encountered:  Today’s telecommunications customers are increasingly embracing digital services and media, and are not only demanding more uptime and quality of service from their service providers, but also instantaneous responsiveness. Swisscom, which operates in Switzerland and Italy, develops, wants to give them just that.

The company, which produces and markets network infrastructures and related services for the telecommunications, information technology, broadcasting, media, and entertainment industry, recognized that its market is shifting beneath its feet as digitization takes a hold in the country. The company had a virtual monopoly inside Switzerland but recognized that new suppliers and players were increasingly disrupting its market. The Swiss telecom market was becoming competitive. Executives recognized this was no time for complacency and needed to be responsive to current and prospective customers as they visited or engaged with the company.

Meeting the challenge: Personalization is the key to superior customer experience in today’s economy. Working with Adobe Experience Cloud solutions, Swisscom developed a platform to serve customers with highly tailored digital experiences, as well as help its employees better understand customer behavior.

See also: Case study — Building a smart car, smartly

A key challenge in Swisscom’s market is the fact that customers speak four languages. Switzerland itself has three official languages — French, Italian and German. Swisscom employed a front-end content management system for its website that communicates with German, French, or Italian, or English-speaking customers who may be browsing on many different devices. Once customers arrive at the website, analytics tracks important data, including number of visitors, sales, downloads, and uploads, which can then be leveraged directly for AB testing. Swisscom’s first AB tests were primarily aimed at measuring customer behavior data to support internal decisions, such as which colors to avoid on banners and where calls to action should be placed. These regular tests, the case study reports, were extremely successful and resulted in an average uplift of 40%.

But to keep up with the fast pace of change, Swisscom, which already employs 17,000 people, couldn’t rely on human employees alone. The company employed Adobe Sensei, an artificial intelligence and machine learning framework, to expand its capacity to personalize and optimize customer experiences. In addition, the analytics generated from the system helps decision-makers better understand visitor behavior, through analysis of number of visitors, downloads and sales across the portal. As a Swissom executive put it, the company started out with AB testing to find out what works best for customers, and now goes beyond AB testing to employ machine learning to deliver relevant news and experiences to a wide range of customers.

Benefits From This Initiative: As a result of its AI-enhanced customer service, Swisscom is automatically improving returns without the need to manually monitor and adjust performance. In addition, data is automatically sent to next-generation testing capabilities, enabling the telecom provider to improve returns and target audiences even while testing experiences. Unlike manual testing and targeting methods, the AI-based system automatically adjusts performance, becoming more effective as time goes on.

The system also provides personalized recommendations that create cross-sell and upsell opportunities and personalizes customer experiences. The results from the targeted approach have been extremely strong, with uplifts of up to 500% compared to manual approaches, the company reports.

(Source: Adobe)

Leave a Reply

Your email address will not be published. Required fields are marked *