What Is Personalisation in eCommerce and Marketing?
Personalisation tailors content, offers, and experiences to individual customers using data. Learn the spectrum from basic to advanced.
Key Takeaways
- Personalisation uses customer data to make experiences more relevant to individuals
- Rule-based personalisation uses segments; algorithmic uses machine learning
- Email personalisation is the most accessible starting point
- Irrelevant personalisation is worse than no personalisation
What personalisation is
Personalisation is using customer data to tailor content, products, offers, and experiences to be more relevant to individual customers. At its simplest, it is addressing an email by name and recommending products based on past purchase history. At its most sophisticated, it is a website that dynamically changes layout and content in real time based on individual behaviour.
Rule-based vs algorithmic
Rule-based personalisation uses predefined segments: if the customer is in Segment A, show creative X. It is easy to implement but limited in granularity. Algorithmic personalisation uses machine learning to optimise at an individual level — Netflix's content recommendations, Amazon's product recommendations. It is more powerful but requires significant data volume and technical investment.
The personalisation spectrum
Most businesses start simply: first-name personalisation in emails, browse and purchase-based product recommendations, location-based content. Moving up: segment-specific email content based on purchase history, dynamic website content based on traffic source or behaviour. The upper end: individual-level algorithmic recommendations, real-time pricing personalisation.
Where to start
The highest ROI personalisation investments for most eCommerce businesses are: first-purchase follow-up email personalised to the product bought; browse abandonment email showing the specific products viewed; cart abandonment email showing the specific cart contents; and post-purchase recommendation email based on purchase category.
The failure modes
Bad personalisation is worse than none. Calling a customer by the wrong name destroys trust. Recommending a product they already bought shows inattention. Targeting with wrong-assumption content is embarrassing. Every personalisation implementation needs data quality checks and creative review to avoid these failure modes.