How should companies judge and punish *us*?
Signing companies up to give their customers clear explanations, use gradation in punishment, do contextual review of a user's full history, and share data on volumes and outcomes
Customers are angry.
Marc Andreessen1 has caused a stir recently by alleging on Joe Rogan’s podcast that many crypto companies and founders have been debanked and had financial accounts shut down in a politically motivated way.
But zoom out from crypto complaints. Debanking is a deeper problem than concerns around crypto or even political targeting, and arises out of the challenges and pressures bank face to punish their customers. And in that, banks are not unique as other companies punish customers too, and the debanked are not alone in being poorly treated by corporate punishment.
Companies increasingly punish their customers at massive scale, in disorienting and opaque ways, and to devastating consequences. What we need is a better vision for how all companies should judge and punish us.
Corporate punishment has gone haywire all over
Companies have always punished their customers, and we’re not even talking about Bank of America’s current 0.01% savings rate on customer accounts. Casinos famously kick unwelcome customers to the curb (and online casinos now bust out repeat winners), and bars and clubs have 86ed guests as long as there have been bouncers. In your neighborhood, there’s a doctor who’s banned a disruptive patient. And there’s always been an element of mad libs style viral stories of corporate punishment that you’ll run across of the puzzling (“Disney obsessed couple sues to get back in exclusive Club 33 fan retreat”) or petty (yes, James Dolan is still banning any lawyer who works at a firm suing his properties from entering Madison Square Garden)
But in our electronic age of dominant platforms, the sting of corporate punishment can cut even deeper. Though platform punishment is most often discussed around social media (and here’s a disorienting recent experience from Bluesky), off-the-rails corporate punishment is far from limited to it. Within banks, whether you’re a Muslim fundraising organization or Melania Trump or involved in one of the millions of suspicious activity reports that banks file each year, thousands and thousands of customers have their financial accounts permanently closed annually with no explanation why nor meaningful way to appeal. Or it’s the New York Times writing about how sometimes your child’s online activity leads Google to lock you out of all your accounts and delete their data. Back on Twitter you run across desperate posts about how the dating app Hinge banned you (again!) for impersonating yourself that rack up 7 million views, or pleas for help after eBay bans you after 20 years on the site for selling an antique pill press. Or it’s Airbnb banning you for a 9-year old possession and DUI charge.
To the extent there’s any discussion of how corporate punishment of customers and users should work, it’s almost always only been pressure to punish more or individualized appeals for forgiveness. There’s a need for a different general vision about how corporate punishment should work better.
The current constraints
Of course, companies need to punish some of their customers. The customer may be king, but he may also be a tyrant, and keeping other customers, employees, and profits safe requires punishing and deterring misbehavior.
Corporate power to punish isn’t unchecked. Historically, the hard-fought limit on corporate discretion has been anti-discrimination laws so that companies can’t punish customers based on certain protected categories, like race. And then around the same time in the 1960s as the welfare state expanded, there grew the recognition that beneficiaries needed more protections than that, and Charles Reich’s landmark article The New Property called for and helped usher in greater procedural safeguards in how powerful governmental agencies (and, later, the companies they contracted with) administered, judged, or deprived individuals of these new kinds of property of government benefits and entitlements.
But extending legal procedures from the welfare state to the problem of how companies, especially ones that are vital or dominant platforms, should judge us is the wrong sentence to hand down. Corporate flexibility to innovate around punishment process and transparency about how that works is crucial, and companies and their customers are in many different types of relationships. But we need some broadly applicable, meaningful principles for how corporations should better judge and punish their customers.
Clear, specific notice
First, all corporate punishments should include specific information about why a customer is being punished. Partly this would involve undoing some laws that sometimes prohibit that disclosure, but mostly it’s about demanding the wisdom long embodied in the writ of habeas corpus that knowing the specific grounds for detention is a precondition for any further meaningful process protections.
There are goals served by secret or generic rationales. It can make it more comfortable to report without fear of retaliation, though to the extent the bad behavior is clear to the punished even without explanation, they still know where to retaliate. But there are also more cynical benefits, with unexplained decisions simplifying the work of corporate justice and making it disorienting to appeal and difficult to do so successfully. The shield of “privacy” can be wielded as an all-too convenient mute button by companies to never have to explain anything. There’s a lot lost in that silence, including abandoning a long-held goal of punishment to help the punished understand their fault and behave better in the future.
Gradation in punishment
The aspiration to help the punished behave better is frequently forgotten in corporate punishment because companies are often banning customers for life. That’s so common (and so alarming) that hackers impersonate those notices to phish us. Why do companies hand out so many lifetime bans?
It seems partly a legacy of how business justice was previously administered where a disfavored customer would be kicked out and banned from the restaurant or casino, photo taped to a wall somewhere, expiration dates too confusing or unsatisfying to attach to the sentence. And the pattern in casinos isn’t unique: corporate punishment’s inclination towards opaque, infrequent severity mirrors the similar tendency in criminal justice.
And as you read proud contemporary declarations of a “zero-tolerance” approach you can hear the hoped for ring of seriousness and toughness in a company’s stance. It’s only later when judgments go wrong that you realize that zero-tolerance punishment isn’t so much a theory of just sanction as a massive claim about the reliability of any process to not mistakenly detect a single fault. But ultimately, lifetime bans are what we as customers too often look for, the assurance of permanency, the unease at surfing in a platform granting a second chance to someone who acted badly before, and the PR risk we then create for companies of not locking someone out and throwing away the username keys.
Another model may be possible. One of the lessons of criminal justice research and reform is that increasing punishment length has little impact on improving ex-ante behavior but large social costs. Pushing platforms to shorter, more certain punishments – months, not lifetimes – and especially for new offenders, may preserve almost all the deterrence, cool down, and corrective goals of punishment, if we’ll allow companies to do it.
Contextual review
If informative rationales and gradation in punishment harken back to the best of legal wisdom, a third principle for corporate punishment to judge users holistically and less by specific incidents is a valuable departure from criminal legal process. Our legal system understandably judges and largely punishes looking just at individual acts. And companies (and their occasional regulators) seem to have adopted that orientation in their trust, safety, and compliance work. But corporate justice, especially on platforms we regularly engage with, has other options.
Large tech-enabled companies focus intensely on getting to know their users in hyperspecific detail, all our interactions – the likes and messages and transactions adding up over years – continually evolving how we are served and what we are offered. Corporate justice should aspire to similarly incorporate a user’s prior history when updating their view of a customer in response to an instance of potential bad behavior. Maybe paradoxically, that will require more help from algorithmic pictures of user behavior. But it could get corporate justice closer to the holistic and probabilistic way we interact with each other as coworkers, repeat customers, friends, and romantic partners.
Public transparency on volumes and outcomes
The debate about corporate punishment is impoverished because we know so little collectively about how it works. In addition to how corporate punishment should work for an individual customer through informative rationales, gradation in punishment, and holistic judgment, large companies should also disclose the outcomes of their overall punishment processes. How many customers were punished, what were the broad categories of reasons they were punished, and how long were they punished? How many customers appealed, how long did that take, and how many won those appeals? Almost no company shares any of this data. That information can broaden and ground discussion – now mostly hidden in anonymous long forum posts or unsuccessful business complaints, and yes, the occasional leading podcast appearance – about how companies are policing activity, whether their initial rulings are well calibrated, and if their appeals review is independent or an electronic rubber stamp.
Informative rationales, incremental punishments, holistic judgment, and data transparency about the number and type of punishments and appeals: some companies’ justice systems have bits and pieces of the above. But as companies and platforms exert a growing influence on our possibilities for personal and commercial exchange, we should sign up the process of corporate justice for more.
His firm is an investor in my employer, and of course all opinions here are my own.