Personalization is no longer optional.
Customers expect brands to recognize context, remember preferences, and reduce friction. At the same time, regulators are applying more scrutiny, and customers are becoming more sensitive to how their data is used.
That tension is where many personalization efforts quietly break down.
What teams often interpret as “better targeting” can feel very different on the receiving end. Helpful intent turns into discomfort. Relevance starts to feel like surveillance. And trust erodes long before any metric flags a problem.
The issue usually isn’t a lack of data.
It’s a lack of clarity about what personalization is actually meant to do.
Reframing personalization as a service
The way out of this tension isn’t collecting more data. It’s reframing personalization as a service problem.
At its best, personalization helps a customer do what they came to do:
faster, easier, and with less friction.
At its worst, it surfaces information in ways that surprise people and raise uncomfortable questions about how closely they’re being watched.
That difference doesn’t come down to algorithms.
It comes down to judgment.
A simple way to think about personalization data
The model below captures how I think about personalization when working with business and marketing leaders.
It separates:
- the types of data customers expect you to use
- from the types of data that tend to create disproportionate legal, reputational, and trust risk
This isn’t about eliminating personalization.
It’s about being disciplined about which data earns the right to be used.
Most practical personalization value lives in a small set of low-risk data:
contextual signals, first-party behavior, and zero-party preferences customers intentionally share.
Most risk comes from a much smaller set of high-sensitivity data that customers don’t expect to be surfaced, even if it’s technically available.
Confusing the two is where teams get into trouble.
The deeper explanation
In the video below, I walk through this model in more detail.
I explain:
- why “more data” is usually the wrong instinct
- how teams accidentally cross the line from service into surveillance
- and how to apply a simple rule of thumb when deciding what data to surface in personalization efforts
If you’re responsible for growth, customer experience, or compliance, this is one of those areas where a small shift in how you think about the problem can prevent a lot of downstream damage.
A final thought
Personalization doesn’t fail loudly.
It fails quietly, through subtle moments of discomfort that slowly chip away at trust.
Getting it right isn’t about being clever.
It’s about being respectful, intentional, and clear on what value you’re actually trying to create.