Data-driven segmentation

Use techniques such as clustering to segment customers based on behaviour (example types of products they browse, value and frequency of spend, browsing behaviour, profile information), rather than predetermined business attributes. Address these clusters with a coherent strategy, executing in near real-time via email, media or basically anything that can be activated based on an email address. Start making informed decisions across different customer touch-points. Identify and prioritise groups of consumers you might be missing or neglecting.

Interest-based segmentation

This type of segmentation allows to find people with similar interests. The approach is to:

1. Derive customer interests based on web data and purchase data (for example categories, discount)
2. Use cluster analysis to find advanced groups
3. Find groups that are most similar to each other within each group and distinctive to others between groups.
4. Score new customer on a daily base in Google Big Query
5. Create persona’s to stereotype segments

Integrating External Data

Use weather, location and other external data sources to enrich the customer micro-moments. For example, trigger promotions for certain products based on the weather in a city on a particular day for a particular consumer demographic, incorporating creative messaging that resonates and adds value to the customer. Go a step further and incorporate local happenings, events, topics into the messaging for a more relevant and meaningful communication interaction.