AI personalisation for websites
AI personalisation shows each group of visitors the content that matters to them and shortens the path to an enquiry.
Background
Existing customers see different content from first-time visitors, and buyers see different content from engineers. This pays off where audiences have clearly different questions and the site has enough traffic to measure the differences.
Rules and models
Clear rules are often enough, based on campaign source, region, the industry in a customer account or recently read topics. Machine learning pays off with plenty of content and data. It finds patterns that rules cannot capture, such as signs that someone is ready to buy.
Examples
Home page matched to the visitor's industry or campaign
Topic suggestions based on recently read articles
Offers for logged-in customers matched to their contract or product
Shop notices based on basket contents and stock levels
In B2B, the campaign link alone often reveals industry or interest. Content that shows how much a site knows about someone's behaviour damages trust.
Consent and technology
Personalisation based on user profiles generally needs consent. Every plan therefore includes a version for visitors without consent.
Edge delivery and lazy-loaded blocks protect caching and load times. In Statamic, editors manage variants block by block. Each variant adds work, so we advise a few well-founded ones.
Measuring impact
Each variant is tested against a control group. Only measurably better variants stay live.