How data science helps brands hyper-personalize their customer experience

By | 4 minute read | November 19, 2020

Businesses and brands must deliver relevant and engaging offers and experiences that directly map to consumers’ ever-changing needs. However, there is still a disconnect between company intent and customer demand. A recent Forrester study found that 90% of companies view personalization as critically important to their business strategies, but just 39% of consumers report receiving relevant brand communications and only 41% report receiving valuable offers.

To address this disconnect, companies are increasingly relying on data science to deliver accurate, relevant, and hyper-personalized experiences to their customers by predicting what they want and need before the customers even know it themselves.

Using data to understand customers and predict their needs

Brands have long sought ways to entice customers to spend—many using promotions, sales, and competitive pricing tactics. Delivering effective promotions is very much a result of thoroughly understanding your customers’ needs. In today’s modern digital economy, this means using your data to build a full 360-degree view of your customers and then using those insights to make valid and effective predictions of their needs.

Boots UK, a British health and beauty retailer, lifted incremental spend with tailored promotions for loyalty card customers by extracting insights from transactional data and using it to deliver relevant promotions to customers. Leveraging data on their 15 million Boots Advantage Card customers—data housed on IBM DB2® database software—Boots UK used the visual, drag-and drop capabilities of  IBM SPSS® Modeler (now a part of IBM Watson® Studio Premium for IBM Cloud Pak® for Data) to build predictive models matching  transactions to individual loyalty card customers. This would then determine the best next action for each person based on their unique preferences and purchase history, and help Boots UK develop and deploy highly targeted marketing messages to these customers.

This analytic and prediction-based marketing approach resulted in a 70% increase in tailored messages and a noticeable uptick in incremental spend from loyalty card customers.

By leveraging their customers’ transactional data, Boots was able to develop a holistic understanding of their customers’ needs and then create highly targeted and effective messages to entice them to buy.

Deploying on hybrid cloud to improve customer service while preserving data privacy

Speed and efficiency are two crucial pillars of good customer service. This is why multinational bank Caixa Geral de Depósitos France (CGD France) accelerated its credit-scoring system with a Watson-infused hybrid cloud app to deliver faster loan service to customers.

Teaming up with EVEA Cloud and Fabrick, CGD France used IBM SPSS Modeler (now a part of IBM Watson Studio Premium) to create digital models for its risk-scoring processes and to re-run historical scoring in line with ACPR regulations. EVEA and Fabrick then developed the credit-scoring application on IBM Cloud® that took advantage of both containers and microservices, specifically the scoring-management module microservice that consumes the IBM Watson Machine Learning service through an API call. Using this hybrid cloud architecture, CGD France was able to store its customer data on premises, where it was covered by data anonymization techniques complaint to GDPR, and then use APIs to access the data when it was needed for the credit-scoring solutions. Furthermore, the mobile app they developed with the Angular framework enabled a single Java codebase to deploy to any web, mobile, or desktop app, so salespeople could use tablet computers to deliver faster loan service from any location.

Using data science in a hybrid cloud approach, CGD France is able to deliver faster, more convenient loan service to customers while staying compliant with data privacy regulations— incorporating the new real-time credit scoring functionality without disrupting its existing core banking systems.

Accelerating and scaling AI models with automation to improve customer experiences

Building and deploying the AI models required for true customer insight can be a time-consuming endeavor that only trained data scientists are able to complete. These challenges prevent the real-time response needed to deliver timely and accurate customer service and predictions to succeed in today’s world. But with the right AI-powered tools, these challenges can be overcome.

IBM Watson Studio, for instance, allows users to build, run, and manage AI models at scale across any cloud. Users can simplify data preparation and data cleansing with a graphical flow editor and visual modelling tools, and automate AI model development with AutoAI, resulting in higher-quality models that are effective and compliant. IBM Watson Machine Learning, also integrated within IBM Cloud Pak for Data, accelerates operationalization of AI models—allowing users to deploy models created in IBM Watson Studio, IBM SPSS Modeler, and open source notebooks. It keeps models accurate with learning and retraining, generates APIs for deployment, and continuously tracks and optimizes AI outcomes. Drive better customer experiences with data insights, predictive analytics, and automated AI.

Read the report on 12 AI-enabled use cases to see how companies have transformed their customer experience and personalized offers with AI.

Read Forrester’s Total Economic Impact of IBM Cloud Pak for Data report to find out how IBM Cloud Pak for Data can help your business increase profitability.