Where Did the Service Go?

How American customer service went from a point of national pride to a broken, bot-managed disappointment — and who's really to blame.

When you do, finally, reach a human being at a modern startup-era company, the experience is often no better — just differently broken.

These are typically young people, often in their late teens or early twenties, hired quickly, trained minimally, and managed almost entirely by metrics There is a particular kind of frustration that has become uniquely American in the 2020s — one that starts with a simple problem, leads you to a chat window, bounces you through an AI loop that understands nothing, and ends with a cheerful automated email informing you that your ticket has been "resolved." Nothing was resolved. You are no closer to a solution. But the ticket is closed, and somewhere, a metric improved.

This is the state of customer service in America today. And if you lived through the 1990s or 2000s, you know exactly how staggering that collapse has been.

A Baseline Worth Remembering

Before 2020, calling customer service was not always a joy — but it worked. You reached a human being, usually within a few minutes. That person had been trained. They knew the product. They had scripts, yes, but they also had the authority and the judgment to deviate from those scripts when the situation called for it.

More importantly, they were polite. Not performatively polite — genuinely, professionally courteous. "Good afternoon, thank you for calling. I'm sorry to hear you're having this issue — let me take care of that for you." Simple words. Powerful words. They acknowledged that your time had value, that your frustration was valid, and that the company owed you an effort.

Even when things could not be fully resolved, you left the call feeling heard. Someone had tried. Someone had cared enough to say: I'm sorry. That matters.

"Good afternoon, thank you for calling. I'm sorry to hear you're having this issue — let me take care of that for you." Those twelve words contained a whole philosophy of service. Today, they are nearly extinct.

Enter the VC-Backed Startup — and the Race to the Bottom

Then came the decade of the startup. Billions of dollars flooded into venture-backed companies whose pitch decks were full of "disruption" and "10x growth" and "unit economics" — but conspicuously light on anything resembling a customer service philosophy.

The calculus was simple and ruthless: customer service costs money. AI bots and offshore scripts cost less. So the bots arrived. And they were — and remain — spectacularly bad at their jobs.

Ask a bot about a billing discrepancy on a subscription you cancelled three months ago, and it will enthusiastically offer you a link to its FAQ page. Ask it to escalate to a human, and it will ask you to rephrase your question. Ask again, and it will apologize for any confusion and offer you the FAQ page again. The loop is perfect in the way that a trap is perfect. It is designed not to solve your problem, but to exhaust your will to pursue it.

Studies consistently show that AI-powered chat tools in customer service fail to adequately resolve the majority of real-world inquiries — the ones that fall outside tidy predefined categories. Which is to say: most of them. The bots handle the trivial. The difficult, the nuanced, the genuinely human problems — those fall into a void.

The bots handle the trivial. The difficult, the nuanced, the genuinely human problems — those fall into a void.

The Human Agents Who Replaced the Humans

: tickets closed per hour, average handle time, customer satisfaction scores. They are evaluated not on whether your problem was actually solved, but on whether the ticket was moved out of the queue.

This is not an indictment of the young people themselves. Many of them are trying their best with the tools and training they have been given. The fault lies with the companies that handed them a six-week onboarding and a scripted response library and declared them ready to face the full complexity of human problems.

Because here is what customer service actually requires: life experience. The ability to hear frustration and understand its root. The judgment to know when a rule should bend. The empathy to recognize that when someone is upset about a $47 charge, they are not just upset about $47 — they feel disrespected, unheard, like they don't matter. A twenty-two-year-old who has never paid rent, navigated a dispute with an insurance company, or fought to get a refund from an airline does not instinctively understand that. It is not their fault. It simply takes time to accumulate that kind of understanding.

But the companies do not care about that. They care about throughput.

Close the Ticket. Declare Victory.

The most maddening development in modern customer service is the premature ticket closure. You describe a complex problem. The agent — human or bot — offers a response that addresses approximately fifteen percent of what you said. Then the ticket is marked "Resolved."

You write back to say it is not resolved. Sometimes a new ticket is opened. Sometimes you are simply ignored. Often, you receive an automated survey asking how satisfied you are with the resolution of your issue, which is its own special kind of insult.

This practice exists because closing tickets looks like productivity. The metrics say: issue resolved. The dashboard is green. No one in the chain of command is asking whether the customer's problem actually went away — because that would require a different kind of measurement, a harder kind, one that valued outcomes over activity.

The dashboard is green. The ticket is closed. The problem remains. Somewhere, a KPI improved.

The Disappearance of Manners

Perhaps what stings most about the decline of customer service is not the incompetence — it is the rudeness. Not aggressive rudeness, usually, but the quiet, careless kind that comes from never having been taught otherwise.

Please. Thank you. I'm sorry — it was my mistake. I apologize for the inconvenience. These are not just social niceties. In a service context, they are load-bearing phrases. They communicate respect. They acknowledge the power dynamic — that the customer came to you with a need, and you have an obligation to meet it with grace.

Today, these words have largely vanished from customer service interactions. Agents move to the problem without greeting. They offer solutions without apology when the company clearly erred. They close conversations without thanks. Some do not even sign their messages with a name — just "Support Team," as if individual accountability were too dangerous to risk.

When a company makes a mistake — a wrong charge, a missed delivery, a broken product — the correct response is an apology. Not a defensive explanation. Not a redirect to the terms of service. An apology. "I am sorry. This was our error. Here is how we will fix it." That sentence, or some version of it, used to be common. Now it feels almost radical.

Somewhere along the line, legal teams convinced companies that an apology was an admission of liability. And so the apology died. And with it, a little more of the relationship between business and customer.

Why This Matters Beyond the Inconvenience

You might be thinking: this is just customer service. It is an annoyance, not a crisis. But I would argue it is a symptom of something larger — a broad devaluation of the idea that businesses owe their customers something beyond the minimum necessary to complete a transaction.

Customer service, at its best, is a statement of values. It says: we want your business, we respect your time, and when we fall short, we will own it and make it right. That statement, made consistently over years, builds trust. And trust, compounded over time, is the foundation of any durable business.

The venture-backed startup model largely rejects this. It optimizes for growth metrics, not loyalty. It acquires customers through advertising and loses them through indifference, and as long as acquisition outpaces churn, the model works — until, suddenly, it doesn't.

But the customers remember. The experience of being brushed off by a bot, ignored by an undertrained agent, and handed a closed ticket that resolved nothing — that experience does not leave people. It accumulates. It shapes how we talk about companies to our friends, how we choose to spend our money, how much goodwill we extend the next time something goes wrong.

What Good Still Looks Like

It is worth noting that not all customer service has collapsed. There are still companies — some large, some small — that have held the line. Companies where calls are answered by people who know the product, who apologize when they should, who say please and thank you and mean it. They are increasingly notable precisely because they are increasingly rare.

They tend to share certain qualities: genuine investment in employee training, longer average tenure in customer-facing roles, and leadership that treats service not as a cost center but as a competitive advantage. These companies understand something that the spreadsheet-optimizers do not: that the way you treat someone when something goes wrong tells them everything they need to know about whether you deserve their loyalty.

A Final Word

Customer service is not a glamorous subject. No one is writing thinkpieces about hold music or ticket resolution rates. But it is one of the most direct, daily expressions of how a business values — or does not value — the people who make its existence possible.

We have lost something real in the past several years. We have lost the competent, courteous professional on the other end of the line. We have lost the sincere apology. We have lost the sense that when you have a problem, the company you're dealing with will genuinely try to help you — not just close the ticket and move on.

It does not have to be this way. The tools and knowledge to do customer service well have not disappeared. The will to do it — and the business model that incentivizes it — is what is missing.

The tools to do customer service well still exist. What's missing is the will — and the business model that rewards it.