A chatbot on an insurance company’s support page opened with my first name last month. “Christian, it’s good to hear from you.” I had typed that name into a form at some point, or it sat in an account profile the company already held, and the software did what it was built to do: pull the field, drop it into the greeting, move on. Ten years ago I would have wondered how they knew. Today I read the same greeting as the opening of a request.
The trick is old. Email newsletters were already running it when I was younger. “Hello Mr. Christian.” “Hello Mr. Christian Huser.” “Hello Huser,” when the merge field caught only the surname. It read as warmth at the time, because so little software bothered. A name in a subject line felt like someone had looked. In fact a database column had matched a row and a template had inserted the value, and nobody had looked at anything. The gap between what the address implied and what actually produced it was wide, but it stayed invisible, so the warmth persisted.
Compare that with a table at a restaurant I visit a few times a year. The waiter remembers my name without a screen in front of him, sometimes a handshake, once an embrace after I had not been in for months. That memory takes him almost no effort in a technical sense; he has met me perhaps a dozen times across several years, among however many hundred other regulars pass through his section in that time. But it is a person choosing to remember one fact about me rather than a system storing every fact about everyone, and that difference is what makes it feel like something. I felt at home before I had sat down.
Set the two side by side and the newsletter looks thin. The name was correct both times; that was never the issue. What differs is that the mechanism behind the newsletter scales without limit and the waiter’s does not. He can only remember so many regulars before the gesture stops meaning anything even to him. Software has no such ceiling. Every customer gets the identical greeting the moment the same field exists in the same database, at whatever volume the company processes.
That absence of a ceiling is the business now on offer, sold under the banner of personalisation “to the next level”. The last ten marketing emails I received all opened with my first name. Every one of them was written, in the sense that matters, weeks before anyone knew I existed as a specific customer. The words around the name were fixed. Only the name was mine. I now read these openings as evidence that the system worked as designed rather than that a company noticed me, and my reaction has shifted along with that reading: where I once felt a flicker of “how do they know my name, when did I sign up, why”, I now feel mostly suspicion, and the suspicion arrives faster each time.
A support chat I used a few weeks ago went further. I had tried, a little sheepishly, to get around a subscription I was supposed to be paying for. In my next conversation with the same assistant, unprompted, it asked whether the billing issue from before had been sorted out. It had not forgotten. I had been caught, by something that does not tire of remembering and never decides, on some ordinary day, that a detail is no longer worth keeping. That is the part that unsettled me more than any name ever did: the fact that nothing about the assistant would let a detail slip.
This is what the software vendors call scaling customer experience. Feeling like a valued customer was already a manufactured good before any of this. The concierge who uses your name, the airline agent who addresses you by title: all of it was designed centrally and delivered identically to everyone who qualified, long before software joined in. What has changed is the material the system now works with. Once a company held a name on file. Now every trace a customer leaves feeds the same greeting: what he searched for, what he almost bought, what he complained about, what he tried to cancel. All of it gets folded in, and it still comes down, in the end, to one email or one line on a login screen.
I felt, reading that follow-up message about my subscription, like a mass-produced good myself. The assistant had done nothing wrong. It had simply been completely right, that fast, about someone it does not actually know, and that precision is what made the gesture feel worth less. The simulation keeps getting better, and each instance of it means less to me than the last one did.
When every company’s email opens with my name, the name tells me nothing. And an assistant that recalls last time’s conversation shows only that it is well built; nobody behind the screen knows me the way the waiter does. Whether that gap will keep mattering to me, or whether I will come to expect the system’s kind of attention instead and stop noticing the difference, I cannot say yet.
A person who forgets an order, who runs late, who has an off day, may start to read as neglect once the system has set the bar at flawless and permanent. Or the same forgetting may become the thing worth seeking out, precisely because no system produces it by accident. Both are plausible outcomes of the same trend, and I do not know which one I am living through. What I can name already is the business underneath either outcome: selling a simulation of being known that is good enough to replace the waiter who simply remembers.