Connected To The Whole World On Facebook, Except To The Company That Connects Us

The world’s largest social network appears to have no functioning customer service. This paradox sheds light on the ambiguous relationship between market scale, business model, and quality of support.

About ten days ago, my personal Facebook account was disabled. The Company told me that I had violated its operating rules – without specifying how or why – and that I had 180 days to appeal. I appealed, and my appeal was immediately rejected. Arbitrariness has no time to waste on bores.

So, I still have no idea what crime I committed, given that my Facebook feed consists essentially of reposts of my Superception publications and, from time to time, photos of Paddy and Potter, our two golden retrievers, who would be deeply upset if they realized that they may have been the cause of this exclusion.

But, one or two days after my appeal was rejected, I received a message from Facebook via WhatsApp informing me that my account had in fact been hacked and had been blocked “to protect me.” This magnanimity, which does Meta credit, does not, however, explain the accusations initially made against me or the rejection of my appeal. At least, though, Paddy and Potter are reassured.

The few attempts I have made since then to log in to my account have all ended with the first message reproduced below.

I therefore set out in search of Facebook’s customer service. And that is when my little story became Kafkaesque. Unless I was particularly shortsighted in my search, there does not seem to be any way to contact a human being. Our only support, as Facebook users, is Meta AI, the Group’s artificial intelligence. Full of good intentions – my apologies for this excess of anthropomorphizing -, it told me the procedure to follow to recover a hacked account. Said procedure ended in exactly the same way – surprise, surprise – with the first message below, now the only interaction Meta is willing to grant me.

Or not. When I reached out to Facebook’s official account on Threads, I received no response, as Meta, unlike the overwhelming majority of companies, does not use its microblogging accounts (X, Threads, Bluesky) to resolve its users’ problems.

So Facebook tells me that my account has been hacked, that it is protecting me, and, in doing so, refuses any contact with me and treats me as though I were the culprit. I am blocked, in every sense of the word.

Beyond my insignificant little personal case, this episode raises questions about customer service at corporate behemoths. Does one need to provide quality customer service when one has more than 3 billion monthly active users? Is it even possible?

I will begin by answering the second question, which, paradoxically, is the easier one to address. When one has invested nearly $100 billion in the metaverse, with virtually nothing to show for it, one should be able to allocate a fraction of that amount to provide users, however numerous they may be, with human support. In this regard, I should point out that I defended on Superception the very principle of this investment in the metaverse because, even though I am not in favor of founders having outsized voting rights in corporate governance, I believe that the mission of such founders is to take the risks that other executives do not want to, or cannot, assume. That is why service quality should not be treated as a mere operational matter: It too is a matter of economic trade-off.

But two truths can be relevant at the same time: Mark Zuckerberg may have been right to invest in the metaverse and wrong not to properly treat the users of his platforms. With such a culture, it is hardly surprising that, all else being equal, Facebook’s and then Meta’s forays into the B2B segment have not been crowned with unforgettable success. We are also familiar with the difficulties Google encountered in customer relations as it sought to develop its corporate cloud business.

More broadly, it is obvious that the larger the number of customers served by a company, the greater the challenge represented by the quality of its customer service. But should those customers be penalized for their service provider’s success? The use of artificial intelligence, which represents a form of digital industrialization of customer service, is acceptable as a first level of interaction if it is supplemented, when necessary, by an exchange with a human being. That is, in fact, how artificial intelligence is deployed in most companies: As a mechanism for collecting information, sorting it, and potentially routing people to human support.

By not acting in this way, Meta reveals nothing other than how little importance it attaches to its users and shows that they are not its customers. As the well-known aphorism puts it¹, on major digital services, the user is the product. The real customer is the advertiser. And when one has more than 3 billion products, disqualifying a few of them is not a dramatic business issue. Admittedly, it is more damaging in terms of image, but Facebook has survived far more serious crises in that area than the dissatisfaction of users it unceremoniously casts aside.

And yet, in networks of this size, individual cases are both insignificant to the leaders of a platform and potentially significant: 0.1% of canceled accounts would represent 3 million people. At Facebook’s scale, the exceptions amount to the population of a small country.

Incidentally, that is why I have several times criticized on Superception the tendency of Meta and other major digital companies to communicate alternately in percentages or in volumes, depending on what suits their interests at the time: The fight against a social or security scourge, presented one way or the other, can give the impression of action having a major impact, while that presentation may conceal a far more mixed reality if it is not expressed both in absolute value and as a proportion.

In this case, criticism of Meta’s customer service is far from mine alone. In March 2024, a bipartisan coalition of 41 U.S. attorneys general wrote to the Group to denounce the dramatic increase in account takeovers and account bans on Facebook and Instagram. The letter explains that the attorneys general’s teams are seeing a sustained increase in complaints on this issue, to the point where it is becoming a significant burden for them. The letter also describes the concrete consequences for the people concerned: Hijacking of private messages, scams targeting their contacts, financial risks, blocked business accounts, and so on.

Meta itself acknowledges that its customer service has fallen short: In an article published in late 2025 and updated in March 2026, the Company writes that it “hasn’t always met expectations,” and announces the creation of a centralized hub and an artificial intelligence assistant. That is the one I dealt with. The article emphasizes that “we’ve also improved how to recover your account if you ever lose access with smarter, adaptive processes.” I would not be quite so assertive.

In the end, scale does not excuse the absence of service: There is no reason why the quality of customer service should decline in proportion to the increase in the number of people served. Can we imagine an airline arguing that, because the number of its passengers has increased, it no longer has the resources to take care of those who run into problems at check-in?

—

¹ As Andrew Lewis, aka blue_beetle, famously noted, “If you are not paying for it, you’re not the customer; you’re the product being sold.”