AI Wearables and Customer Data: What Marketers Should Expect

I bought a smart ring partly out of curiosity about what it collects. The answer is a lot: sleep stages, heart rate variability, temperature trends, activity, and increasingly AI generated interpretations of all of it.

That is a category of data marketing has never had. It is worth thinking about now, before anybody builds a business on assumptions about it.

What These Devices Actually Collect

Continuous biometric data, interpreted by a model into something readable. Readiness scores, stress estimates, sleep quality.

The interpretation layer is the new part. Raw heart rate was available before. A daily assessment of whether you are recovered enough to train is a different product.

For health purposes this is genuinely useful. My sleep got better because I could see what alcohol was doing to it, which I already knew and had never seen quantified.

Why Most of This Will Not Reach Marketers

I want to be clear about this, because the speculative version of this article is more exciting and less true.

Health data sits under specific regulatory protection in most markets. Even where a consumer device sidesteps the strictest health rules, the platform operators have strong commercial incentives not to sell biometric data. Their entire premium positioning depends on people trusting the device on their finger.

Apple has made privacy a product feature. Oura and similar companies publish explicit positions on not selling health data. Those positions could change and the reputational cost of changing them is enormous.

The realistic expectation is that biometric data stays inside the device ecosystem.

What Will Reach Marketers

Aggregate behavioral trends. Anonymized, population level insight about when people are active, how sleep patterns shift seasonally. Useful for timing and category research, not for targeting individuals.

Self reported wellness segments. People will increasingly describe themselves using categories these devices taught them. "Poor recovery" and "low readiness" are entering ordinary vocabulary, and that changes how you write copy for health adjacent products.

Contextual moments, mediated by the platform. The device knows you just finished a workout. The platform may allow an ad or an offer at that moment without ever handing over the underlying data.

That last one is the realistic version of the targeting story. You rent the moment, not the data.

The Question Worth Answering Now

If this data did become available, would you use it?

I think a lot of marketers would say yes without considering how it reads to the person on the other end. An ad that references your sleep quality is not clever. It is alarming.

There is a line between relevant and invasive, and it moves faster than most companies' instincts. The cost of crossing it is not a fine. It is a customer who feels surveilled and tells people.

My position on this is the same as my position on data generally, which I laid out in ethical marketing principles. Collect what you need, say what you do with it, do not sell it.

What I Would Actually Prepare For

Nothing technical. This is not a tooling problem yet.

What is worth doing now is deciding your policy before the capability arrives, because the decision is much harder when there is revenue attached to saying yes.

Write down what categories of data your business will not use, even if it becomes legal and available. Companies that decided this in advance handled the last decade of privacy shifts far better than the ones that improvised.

For the regulatory picture, the FTC's health privacy guidance covers where consumer health data sits in the US, and it has been getting more aggressive rather than less.

What Customer Data Should a Small Business Actually Collect?

The minimum that lets you deliver the service and follow up. For most small businesses that is a name, one contact method, what they bought, and when. Everything beyond that needs a specific reason you could explain to the customer without embarrassment.

The test I use with clients: for each field you collect, name the decision it changes. If a field does not change what you do next, it is a liability rather than an asset. It has to be stored, secured, disclosed, and eventually deleted, and it produces nothing.

The second question is retention. Most businesses keep everything forever because deleting feels like losing something. Data you no longer need is pure risk, and a simple policy that removes inactive records after a defined period reduces your exposure at no cost to the business.

Where small businesses most often overstep: buying contact lists, importing contacts from an unrelated context, and adding people to marketing flows because they filled out a support form. All three are common, all three damage trust, and the last one is increasingly a compliance problem rather than an etiquette one.

The practical starting point is an inventory. List every place customer data sits, including the spreadsheet on somebody's laptop and the shared inbox. Most businesses have never done this and are surprised by how many places turn up.

The policy decision is worth making before the capability arrives, because it is far harder once revenue is attached to saying yes. Companies that wrote down what they would not do in advance handled the last decade of privacy shifts considerably better than the ones that improvised under commercial pressure.

The line between relevant and invasive moves faster than most companies' instincts, and the cost of crossing it is not a fine. It is a customer who feels watched and tells people, which is the most expensive kind of marketing failure. Why dishonest marketing is dying makes the general version of that argument.

Would you use biometric data for targeting if you could? I do not think I would, and I am aware that is easy to say while it is hypothetical.