Customer Data Platform for Omnichannel Personalisation

📋 Overview

A national fashion and accessories chain with 45 stores in shopping centres and on the high street, an online store growing 40% a year, and an active social presence. Its loyalty programme has 280,000 members.

🎯 The Challenge

Business context The chain faced the fragmented-data paradox:

• POS data from physical stores, Shopify for e-commerce, Instagram Shopping and the loyalty programme all lived in separate systems
• The same customer behaving differently in each channel was never recognised as one person
• Generic marketing campaigns were converting at a declining rate
• The marketing team depended on IT to extract any customer segment
• No way to measure customer lifetime value or predict churn

Technical requirements

• Identity unification: resolve one person across email, phone, loyalty card and cookie.
• Real time: profile updates in near real time after a purchase or interaction.
• Advanced segmentation: predictive models for purchase propensity and churn risk.
• Self-service marketing: the marketing team builds segments and campaigns without IT.
• GDPR compliance: consent management and right to erasure built in.

💡 The Solution

Strategic approach A phased Customer Data Platform on Azure, starting with data unification and identity resolution, then moving to predictive segmentation and campaign automation.

Technical implementation

• Azure Data Factory: ingestion from POS, Shopify, the loyalty programme and social interactions, on schedule and in near real time.
• Azure Synapse Analytics: a central warehouse holding a golden-record unified customer profile.
• Azure Machine Learning: RFM segmentation (recency, frequency, monetary), purchase propensity by category, and 90-day churn prediction.
• Dynamics 365 Marketing: automated customer journeys driven by the predictive segments, with multichannel campaigns — email, SMS, push — and dynamic content.
• Power BI: customer analytics dashboards — CLV by segment, campaign effectiveness, cohort analysis and multichannel attribution.
• Consent management: an integrated GDPR module with granular consent and a right-to-erasure workflow.

Timeline

• Phase 1 (data unification): 6 weeks — ingestion, cleaning, identity resolution
• Phase 2 (analytics and segmentation): 4 weeks — ML models, dashboards
• Phase 3 (campaign automation): 4 weeks — journeys, personalisation
• Phase 4 (continuous optimisation): ongoing — A/B testing, model refinement

📈 The Results

Success metrics

• 🎯 Conversion: personalised campaigns convert 35% better than the generic campaigns they replaced.
• 💰 CLV: average customer value up 22% in the first year.
• 📉 CAC: cost of acquiring a new customer down 45%.
• 🔄 Retention: churn among loyalty members down 30%.
• 🚀 Marketing autonomy: the marketing team builds segments and launches campaigns without raising an IT ticket.

Business impact The CDP changed the brand’s relationship with its customers: from mass generic campaigns — “20% off everything” — to communications based on real behaviour. Customers receiving personalised offers open at three times the rate and convert substantially better.

🛠️ Technology Stack

• Azure Data Factory: data ingestion and orchestration.
• Azure Synapse Analytics: warehouse and unified customer profiles.
• Azure Machine Learning: predictive segmentation and churn models.
• Dynamics 365 Marketing: multichannel campaign automation.
• Power BI: customer analytics and campaign effectiveness.

💬 Client Testimonial

“We finally know who our customers are — not as numbers, but as people with real preferences. The difference between sending ‘summer arrivals’ to the whole base and ‘the floral dresses you love are 30% off’ is enormous.”

Marketing Director Fashion & Accessories Chain