Yesterday I was looking for shoes, not on a platform or through an app. I told a friend my running shoes were done. This morning my phone shows me running shoes.
It could be coincidence. I have no way to check any more.
The marketing word for what may have happened to me is hyperpersonalisation. The pitch sells it as the next stage of customer experience, content cut precisely to one person, dressed in words like precision and relevance. I was genuinely excited at first.
What precision requires
Precision at that level needs a lot of input. Purchase history, browsing behaviour, location, time of day, dwell time, click patterns, social connections and search history all feed the model. During the ecommerce boom I collected exactly these data points myself, matter-of-factly and cleanly. But those were the early days.
Today the same collection amounts to constant surveillance of a person.
The old line, a free service in exchange for data, undersells what is actually gathered now. It goes well past age and postcode. Hesitation counts too: how long someone looks at an item before scrolling past. So does the time of day a purchase happens, and what gets searched and never bought. The gap between search and purchase tells a company more about a person than the purchase itself does.
There is no honest name for that other than profiling.
Why it feels like the opposite
The elegant part of hyperpersonalisation is that it feels good, and that is the trick. You get what you want before you know you want it. You enjoy the sense that someone understands you, and barely register that you are being watched. Perfect surveillance feels like attention and courtesy.
The concept contradicts itself. Hyperpersonalisation is industrial mass production of uniqueness: millions of customers, each meant to feel individually seen. That illusion survives only if a company knows each of those millions well enough to sustain it, which means watching each of them without a gap.
There is an older name for comprehensive observation paired with steering behaviour on the back of what that observation finds. Shoshana Zuboff wrote the definitive work on it: surveillance capitalism. I don’t think the people selling hyperpersonalisation have read it.
What actually bothers me
The technology itself does not bother me. A doctor who knows my medical history treats me better for knowing it. It is the same mechanism, though in a context I feel I have under control.
What bothers me is the vocabulary. Everywhere I read about customer experience, added value and relevance, and never once the plain word for what is happening. The words skirt surveillance and power on purpose. Least discussed of all is what happens once someone holds this much data on a person and puts it to a use other than recommending shoes.
A customer profile precise enough to show me the right shoe at the right moment can just as easily show someone else the right political message at their most vulnerable moment. The same infrastructure does both; only the use changes.
Every time I read that someone should feel unique, I think about how much the model had to collect about that person for it to work.