Stockertim Dictionary · Artificial intelligence · Business
Personalisation
Also called personalised marketing, marketing personalisation, personalized marketing
In short
Tailoring marketing content or offers to a person or a group of customers based on data about them, such as what they bought or said they prefer.
- Beginner
- Process
- Checked
In detail
Personalisation means not sending everyone the same message. At its simplest it is a name in a greeting; more often it means choosing which products, offers or articles each person sees based on what is known about them, such as past purchases, location, stated preferences or recent browsing.
It can work at two levels. Segment personalisation groups customers into clusters, such as new buyers, regular buyers and lapsed buyers, and sends each group a different version. One-to-one personalisation uses a model to pick content for each individual, for example a recommendation engine that ranks products by predicted interest.
Personalisation depends on personal data, so consent and preferences matter. Data protection law in many countries limits how personal data may be used, and marketing messages often need permission. Rules differ by country. Beyond the law, there is a line between helpful and intrusive: a reminder about an item left in a basket can feel useful, while a message that reveals the business has inferred something sensitive, such as a health condition, can damage trust.
Good practice is to use data people would expect the business to use, let them change their preferences easily and check that segments add up so no group is missed or messaged twice.
Worked example
Harbour Foods splits 12,000 email subscribers who agreed to marketing into four segments for a seasonal campaign.
- Segments: 4,800 + 3,600 + 2,400 + 1,200 = 12,000 (regular, occasional, new, lapsed)
- Regular share: 4,800 / 12,000 = 40%
- Lapsed share: 1,200 / 12,000 = 10%
- Check: each subscriber appears in exactly one segment, so the counts add up to the list.
Four versions of the email go out, every subscriber receives exactly one, and the lapsed group of 1,200 receives a win-back offer instead of the standard promotion.
In practice
Retailers personalise product recommendations and emails, publishers personalise article lists and service firms tailor reminders. Teams document which data each rule uses, honour opt-outs promptly and review messages that rely on inferred traits.
Watch out
- Having data does not mean customers expect or accept its use; intrusive personalisation can lose trust.
- Segments that overlap or leave gaps send some people two messages and others none.
Learn it properly
Official sources
- Artificial intelligence · Information Commissioner's Office
- A guide to the data protection principles · Information Commissioner's Office
Facts last checked on 3 November 2026.
Education, not advice: the Stockertim Dictionary explains what terms mean. It never recommends buying or selling anything.