Personalisation With AI: Showing Visitors What They Need

Personalisation means adapting what a visitor sees based on who they are or what they have done. Done well, it saves people time and increases conversions. Done badly, it feels invasive.

Useful examples

• Product recommendations: suggestions based on browsing or purchase history.
• Content suggestions: “related articles” tuned to a reader’s interests.
• Location-aware information: nearest branch, local delivery times or currency.
• Returning visitor messages: continue where they left off, or show relevant offers.
• Smarter site search: understanding natural-language queries and typos.

Where AI adds value

Machine learning can find patterns in behaviour that rules alone would miss, and language models can improve search, summaries and support. But even simple rules, such as showing different content to new and returning visitors, can deliver strong results.

Respect privacy

• Be transparent about what you collect and why.
• Get consent where required and honour opt-outs.
• Use the minimum data needed.
• Avoid inferring sensitive traits such as health or financial situation.

Do not overdo it

If personalisation feels like being watched, it backfires. Keep the experience useful, predictable and easy to reset.

Start small

Pick one journey, such as the home page for returning visitors, and test a simple personalised element against a standard version. Keep it if it improves the result, and expand carefully.