Personalisation with customer engagement matters. There's no denying it.

People expect more from the brands they trust with their data, from basic privacy through to actually being understood.

High-profile breaches over the years, at major hotel groups and social platforms, have made people far more aware of what data they're sharing, who they're sharing it with, and what they expect in return. Brands need to treat customer and member data with real care, not just a privacy policy nobody reads.

It's not only about breaches, either. Several major tech platforms have openly admitted to using conversation data for ad targeting, and it's the kind of thing that makes people uneasy. We're not suggesting you eavesdrop on your customers. We're suggesting you understand their behaviour, respect the data they've given you, and use it well.

So what do you actually do with the data?

Use it wisely, and wisely means to your member's benefit, because that benefit comes back to you. There's a mountain of research on this, but the short version is that personalisation done well drives real loyalty. Consumers get frustrated by impersonal experiences, they're more likely to become repeat buyers after a personalised one, and they often expect a personalised offer within hours of identifying themselves to a brand.

How much data do you actually have?

Probably more than you'll ever use, but having data and understanding it are two different things. Numbers and dashboards mean nothing until you know what they're actually telling you and what to do about it. Start simple: get the name right instead of "Dear Member," and get the language right if you're operating across markets. It sounds basic, but plenty of brands still get this wrong, whether it's legacy systems or simply not paying attention. Where a member has openly shared information, or where you can reasonably infer something (a title of "Mr" suggests gender, for instance), use it respectfully. It shows the most important human quality a brand can demonstrate, which is that it's actually listening.

Here's a real example.

I once banked with a large bank while living in Singapore. I'd walk into a mall, complete a purchase, and get an SMS offering a discount at a specific store in that mall if I used the same card again. The store was Nine West, a women's shoe and accessories retailer. I'm a guy. I'd never bought from Nine West, never expressed interest in it, and never ticked a box.

• Relevance score: zero out of ten.
• Spam score: nine out of ten.
• Do I think that bank is a bit clueless? Ten out of ten.

Years later, that's still the story that comes to mind whenever someone mentions that bank.

A few rules worth keeping close: don't be creepy, since nobody wants their friends or family to know what they've been buying. Don't use every variable at once, because each one is another chance for something to go wrong. Give to get, meaning if you need more information, offer something real in return for it. Listen and act on stated preferences, because nothing erodes trust faster than being asked for information and then having it ignored. Segment dynamically, since people change and shouldn't be left in "Segment A" forever. And test, learn, and do it again, because member relationships evolve and your approach should too.

Don't be a "Dear Bob" brand. Context is everything.