Why does the human customer service only appear after I click "I want to complain"?

Ask AI · How does the abuse of complaint mechanisms affect the rights of ordinary employees?

Do you know the fastest way to reach human customer service?

Undeniably, the introduction of AI is driving the reshaping of customer service roles. In some basic customer service tasks, AI is undoubtedly the “main force” now, but in actual experience, users’ needs vary widely, and AI customer service often cannot provide perfect solutions.

Moreover, we are gradually discovering that now it’s really not easy to find human customer service.

Some people are “looped around” by AI for 30 minutes, finally entering the waiting queue;

Some have shouted “human customer service” 10 times, and finally found a person;

Some listen back and forth to what “#1-#” buttons mean three times, confirming there’s really no “human customer service”…

I don’t know when it started, but “complaints” and “reports” have become the most direct and quick ways to find human customer service. And as more and more people use the “complaint” function, new problems have emerged: every useless call and every malicious ticket occupy channels that should be for those truly in need of help.

So, do you think “complaints” are being abused?

Some positions simply can’t withstand even one “complaint”

@Today I’m drinking oolong tea

After calling the courier company’s customer service three times in a row without being able to reach a real person, I helplessly chose to complain.

The reason was that I bought a laptop, and during the waiting process for delivery, I received a message asking me to choose a delivery method. Because it was a valuable electronic product, I wanted to inspect it in person, but there was no “door-to-door delivery” option. So I hoped to communicate with customer service and ask them to give me a way to select “door-to-door delivery.”

Unbeknownst to me, during my search for human customer service, the delivery status changed to “Out for delivery,” and a specific delivery person was assigned. Therefore, when I used “complaint” to reach human customer service, the complaint also landed on this delivery guy.

With the help of customer service, my issue was quickly resolved. During this process, the delivery person also contacted me via work phone, and I clearly stated that I had already resolved the issue through customer service.

The turning point happened a day later.

I first received a strange call. After answering, the delivery guy immediately identified himself and said that because of that “complaint,” he received a suspension notice that day, and the company also asked him to hand over his work phone. The guy not only begged me to call back and withdraw the complaint but also said he could privately send me mooncakes.

Later I found out that because that brand provides dedicated lines for the courier company, my “complaint” caused by “inquiry” was directly transferred to the postal management department overseeing the courier, and the involved delivery guy was also affected. The only way to “save” him was that call to withdraw the complaint.

This story happened before AI customer service was widespread, but it truly impacted me. When an inquiry is upgraded to a complaint, it can “accidentally harm” ordinary people.

Before talking about abuse, let’s first solve the difficulty of rights protection

When we discuss whether “complaints are being abused,” the voice of “difficulty in rights protection” cannot be ignored. Although more and more consumers are familiar with how to proceed with rights protection, stories of missteps are everywhere.

Even experienced rights defenders “lose battles”

@Zuo Xi

This is a story of an experienced rights defender “losing a battle” despite having sufficient evidence and quick responses.

During the Spring Festival this year, my friends and I traveled to the Philippines. Considering the chaotic regulatory system in Southeast Asia, we chose an international hotel chain and booked through a well-known domestic OTA platform.

On the first night, after arriving, we didn’t carefully inspect the room before sleeping. The next morning, we found a dried brown stain on the back of the pillow, which confirmed our suspicion when we searched thoroughly and found a mouse droppings.

As seasoned rights defenders, we quickly started the process of collecting fixed evidence: taking photos, videos, close-ups of the mouse droppings, one person staying in the room to watch the evidence and prevent it from being taken away by cleaning staff, and another going downstairs to negotiate with the hotel.

Our demands were to change rooms and refund the first night’s fee. But what made us furious was that the other side not only didn’t apologize but also said they could only change rooms, not refund, claiming “the guest didn’t find the mouse during check-in.”

“Did you see the mouse with your own eyes?”

“You didn’t see it yourself, how can you prove there was a mouse?”

Throughout the negotiation, they kept arguing that “if you didn’t see the mouse, then there was no mouse,” and in the end, they even fell silent and refused to communicate.

Considering we still had two days before check-out, and the order was not yet completed, the money was still held in the platform’s escrow account. As the transaction platform and operational agent, the OTA platform has a responsibility to protect customer rights. Coupled with the convenience of communication, we chose to negotiate with the OTA platform.

Unexpectedly, talking to customer service was just the beginning of a nightmare. “We understand your feelings… We are actively communicating with the hotel… Please wait a moment…” The online customer service’s responses sounded like a broken record. If it weren’t for the employee ID in front, I would have suspected AI was answering.

Based on previous rights protection experiences, we explicitly asked the customer service to give an accurate reply and solution by 6 pm that day. When the time came, the platform’s customer service appeared as promised but told me they “could not contact the hotel.” Despite my anger, I remained patient and told the customer service, “If you can’t handle it, transfer me to your superior department, and give me a result by tomorrow.”

The next day, the customer service reappeared with the message “The hotel refused to refund, and the platform is powerless.” With evidence in hand and the order still incomplete, the so-called “powerless” claim from a well-known domestic OTA platform was hard to believe. Since I couldn’t call the consumer association abroad, I could only communicate with the platform online.

On the third day, the customer service brought a very perfunctory solution: “The hotel refused to refund, and the platform will compensate 80 yuan.” The double inaction of the hotel and platform deeply stimulated us. Throughout the trip, we kept negotiating with the platform, and the final solution was “hotel refused refund, platform compensated 80 yuan, agent compensated 100 yuan.”

After returning home, we called the 12315 consumer hotline to continue holding the platform accountable, initiating nearly a month of “passing the buck.” The initially promised 180 yuan compensation was nowhere to be seen; the platform claimed it had paid, but no record appeared in the transaction history.

During the accountability process, we unexpectedly discovered that the actual order was made on an overseas OTA platform, which the domestic OTA had concealed from the beginning, and failed to communicate promptly, causing us to miss the golden window for in-store rights protection.

To this day, our rights protection is ongoing, and the consumer association has not provided effective help. It makes one wonder: we’ve done everything an ordinary person can do, yet still cannot defend our basic rights. If it were middle-aged or elderly people unfamiliar with the rights protection process, would they really just have to silently accept “being wronged”?

Drinking “screwdriver yogurt” but being asked to rate “the food is fine”

@Ms. Tian

The story begins with a delivery yogurt that contained a screw.

As soon as it happened, I took photos to preserve evidence and sent them to the merchant, but the store claimed “there are no screws in the food preparation area,” avoiding responsibility and suggesting I request a refund. Later, I contacted the delivery platform for intervention. Two days later, the platform told me the resolution was a compensation of 57 yuan.

According to Article 148 of the Food Safety Law of the People’s Republic of China, when consumers purchase food that does not meet food safety standards, they have the right to request compensation from producers or operators, with the amount increased by up to 1,000 yuan if the compensation is less than that.

The platform’s handling clearly did not comply with relevant regulations. I immediately cited the relevant clauses to object. The platform replied that they had no authority to punish, so they couldn’t handle it. I understood and then called 12315 for further resolution.

About seven working days later, I received a reply from the relevant department, informing me that my case was not accepted because the merchant held valid certificates, and no violations of food safety procedures or foreign objects like screws were found on-site.

I was dissatisfied with their handling, which only checked the store and didn’t review the surveillance footage from that time. Even more ridiculous, after I submitted a report to the authorities, the shop owner contacted me on his personal phone, admitting to mistakes and hoping I would acknowledge “that I didn’t see clearly myself” on the platform.

With the mindset that “today’s me might be tomorrow’s everyone,” I filed a further complaint with 12315 and requested administrative reconsideration. As of the time of writing, there has been no update.

Reflecting on the entire complaint process, my biggest feeling is that I was negotiating with a “desire to settle” attitude, not wanting to escalate the matter, but every step seemed to ignite the fire of the infringing party’s emotions.

This reflects the current difficulty of rights protection. It’s hard to imagine what will happen if AI fully takes over customer service in the future.

AI customer service, a “good employee” that doesn’t earn wages but blocks rights protection

@Wen Luo

Young people are often confused by AI customer service—what about middle-aged and elderly people?

That day, my mom told me she went out and came back, saying that when she paid the parking fee via her phone, the money was deducted but the parking fee wasn’t paid. I checked her phone and found she was misled by a prominent “discounted insurance purchase” entry on the payment page.

The interface was extremely confusing; the button’s color and position looked very similar to normal payment options, misleading my mom into clicking, and she accidentally bought a travel safety insurance worth dozens of yuan. The money was deducted, but the parking fee wasn’t paid, and she ended up with a policy she didn’t need or understand.

I immediately wanted to help her get a refund, but the refund entry was hidden deep within the page. After finally clicking in, the system coldly prompted: “Artificial customer service is outside service hours, please use intelligent customer service.”

I patiently interacted with the AI, only to find it could only provide policy inquiries, clause explanations, and other self-service guidance—no cancellation or refund options. Experienced in these tricks, I repeatedly entered “refund” and “cancellation,” but the only options that appeared were “cancel service fee,” just a few yuan, clearly telling you: don’t expect to cancel the policy or get your money back.

The next day, during working hours, I called the customer service line. After waiting more than ten minutes, no one answered. Either no one was there, or AI was blocking the line, but this is just the normal after-sales channel.

Helpless, I could only escalate to the platform’s complaint system. Unsurprisingly, the first interaction was with AI. After repeatedly emphasizing “elderly accidentally triggered,” “induced purchase,” “rights protection,” and other strong words, they finally transferred me to a real human agent. After a lot of trouble, I got a full refund.

Thinking about it afterward, it’s especially ironic: a normal mistaken purchase and reasonable refund request couldn’t be resolved through the normal process. The AI customer service was useless, and the human customer service was unreachable. The only way to resolve the issue was to use “complaint” as a weapon, turning complaints into the only shortcut to find a real person.

When normal rights protection requires complaints to advance, and finding human customer service involves a fight, are we making things easier or harder?

—

For ordinary people, customer service is everywhere. They handle not only complaints but also pre-sales inquiries, during-sales assistance, and after-sales service—serving as the “last mile” of the service industry. But the reality is, the still “immature” AI customer service and the gradually “disappearing” human customer service are perhaps both the “last five hundred meters” in that “last mile.”

Finally, I want to share a comforting story.

After waiting on the phone for an hour, I finally connected with a human agent. After explaining the issues with data migration on my phone, the customer service immediately said they would guide me step by step. The process involved long waiting times, but throughout the call, the agent kept comforting me:

“It’s okay, take your time, we’ll stay online, and all problems can be solved.”

Hope everyone can meet good customer service.