Last checked: 7 November 2026
In short: AI personalisation is the use of AI to tailor what each customer sees, such as products, messages or advice, based on that person’s own behaviour and context rather than a broad group they belong to.
Table of Contents
How it works
- Collect signals. The system records what a customer views, buys and searches for, and where they are.
- Predict. An AI model works out what that person is likely to want next.
- Tailor. The website, app or email changes for each person, often as they use it.
Where you’ll meet it
Recommendations on Netflix, Spotify and Amazon are personalised. ChatGPT and Gemini can tailor replies using a memory of past chats.
What it means for your business
- More relevant offers can lift sales. Test personalised pages and emails against standard ones to measure the effect.
- Customer data drives it. Clean data that customers agreed you could use matters more than the choice of tool.
- Start with one channel. Email or product recommendations are easy places to begin and to measure.
Watch out for
- Privacy rules. UK and EU data protection law requires a lawful reason and clear notices before you use personal data this way.
- Unsettling results. Targeting that seems to know too much can make customers uneasy and damage trust.
Example
Zalando. In 2023 the online fashion retailer launched an AI assistant, built on its own models and OpenAI’s language models, in four markets. On 1 October 2024 it opened the assistant to logged-in customers in all 25 of its markets, in local languages. The assistant gives personal fashion advice and takes account of location, weather and occasion.
AI personalisation vs segmentation
| AI personalisation | Segmentation | |
|---|---|---|
| Unit | Each individual customer | Groups of similar customers |
| Updates | All the time, as behaviour changes | Now and then, when groups are redrawn |
| How they relate | Can use segments as a starting point | The older, broader method |
Frequently asked questions
Do customers want personalised experiences?
Many do when it saves time and feels relevant. Trust drops when firms use data customers did not expect them to use, so explain what you collect and why.
Can small online shops use AI personalisation?
Yes. Most online shop and email platforms include recommendation and personalisation features in their standard plans, so small shops can start without building anything.
Is personalised pricing legal?
It can be, but rules on fairness, discrimination and consumer protection apply. Take legal advice before charging different customers different prices based on their data.