Here is a wild story:

You should read the original thread of tweets here, but here’s what happened: 

Pim Techamuanvivit, the owner and chef of a few popular restaurants in San Francisco, was managing the floor at Kin Khao the other night when a call came in from someone asking about their delivery order. This was surprising to her, since her restaurant doesn’t do delivery – not even takeout. After hanging up the phone, she googled “Kin Khao delivery” and found something astonishing: a complete impersonation of their menu and brand, complete with delivery ordering, on Seamless, Grubhub and Yelp. 

Now, there are a couple potential explanations for this; the simplest is that Kin Khao had a profile automatically made for them on these delivery services that shouldn’t have been there (since they don’t do delivery), and that’s the end of the story*. But that’s not my impression of what happened! A real order got placed, and they only found out about it because the customer called them directly. That’s what’s fishy here.

They’re a popular restaurant; they’ll be getting orders all the time. If Pim was never aware of this (and she would know!), then either there are a pile of unfulfilled orders that got ghosted on, and she would have seen couriers come into the restaurant regularly. Or… maybe they were being fulfilled. And had that customer never called Kin Khao directly, no one would ever have known.

Anyway, because this is my newsletter and not actual reporting, I’m going to assume that the most interesting of the scenarios happened, which is: someone made a counterfeit restaurant! This is a whole step beyond the recent allegations that Grubhub and Seamless were putting up landing pages in order to intercept orders and collect commission fees. This isn’t impersonation in order to collect a referral; it’s straight up counterfeiting the restaurant on a delivery app. Unsuspecting customers, recognizing a business that they know, order food expecting it to, you know, actually come from that establishment! 35 minutes later, food arrives at their door. Maybe you notice that the food isn’t quite what you expected, or isn’t up to par. But maybe you don’t notice! If you think, “that would never happen to me”, well, are you sure? 

As the food delivery wars heat up, most of the discussion I’ve heard has been around the food delivery apps’ pros, cons and comparative advantages as aggregators: is demand elastic, who has pricing power, how do they commoditize restaurant supply, and those sort of themes. All aggregators come with their share of societal problems, and food delivery is no exception. 

But there’s another side of the story, and another suite of problems, that I’ve heard less chatter about but long term is way more important. their impact and unexpected consequences as platforms. What new kinds of bad behaviour, like counterfeiting food, are here to stay? What are we going to have to regulate, and who will do it? 

Counterfeit food, of course, isn’t a new thing at all – or at least, counterfeit ingredients aren’t a new thing. There’s olive oil fraud, where cheaper olives from inferior regions or lower-quality processing plants are falsely labeled as extra virgin. You may have heard of “honey laundering”, when cheaper Chinese honey gets mixed in and sold as North American honey in order to evade tariffs. Similar story with the Great Canadian Maple Syrup Heist of 2016. Premium Arabica coffee beans are routinely cut with cheaper Robusto ones. “Kobe Beef” in North America is just a marketing term; so is Prosciutto di Parma. “Parmagiano Reggiano” is even worse: you might be eating sawdust. 

The biggest food fraud category? Seafood. A 2016 report from Oceana outlined just how bad the seafood industry has become. Vast quantities of fish illegally caught, laundered and resold in black markets; up to 20% of the fish we eat can be the wrong species entirely. As summarized on Eater: 

A couple particularly egregious examples of fraud cited by Oceana include bluefin tuna in Brussels restaurants, where 98 percent of the dishes tested actually contained another fish entirely; and a sushi restaurant in Santa Monica, California that was busted serving endangered whale meat as fatty tuna. A species called Pangasius, or Asian catfish, is particularly popular with fraudsters, and has been discovered standing in for 18 different species of fish including cod, flounder, grouper, sole, and red snapper. Another food that’s highly susceptible to fraud: caviar. In one fraud study, 10 out of 27 caviar samples were found to be mislabeled, and three of them didn’t even contain animal DNA but rather a completely unidentified substance.

This stinks, but it isn’t the most surprising thing I’ve ever heard. It’s a natural consequence of the food we eat coming from farther and farther away, with a complex, opaque supply chain abstracted out of sight and often outside the law. Food got abstracted away into the “food cloud”. And these days, we’re adding another abstraction layer on top of that: we’re outsourcing food preparation too: at every income level, for every meal, we’re eating more food that’s been prepared by someone else: cooking as a service. 

As food got abstracted into the food cloud, and now cooking is getting abstracted into the cooking cloud, it should not surprise anyone that you’re not always getting what you think when you hit that order button. Upton Sinclair’s famous book The Jungle, on the Chicago meatpacking plants at the turn of the 20th century, took us inside the food cloud where the sausage gets literally made. It’s not pretty. A century later, Anthony Bourdain’s Kitchen Confidential painted a picture of what really goes on in restaurant kitchens. 

The “integrity crisis of food”, if you want to call it that, is on the one hand to be totally expected. Why wouldn’t food get counterfeited, just like every other consumable good? But on the other hand, it’s a bigger problem than knockoff Nikes. Illegal ingredients are often environmentally devastating, and carry unknown health risks while we’re at it. Knockoff or unlicensed kitchens, if they can operate under the cover of someone else’s health certification, may be cutting food prep and safety corners that aren’t acceptable – plus, they’re outright theft from restaurant owners, stealing their brand while sticking them with the food safety risk. 

But the food cloud keeps growing. Every time we add a layer to the food cloud, we add a new layer of opacity, and a new set of costs that are now subjected to margin pressure. In the old days, you knew exactly what food you were eating, because you or someone near you grew it themselves. But then, for most of us, food provenance got abstracted away into a grocery store: you didn’t know where a stick of butter came from anywhere, it just showed up. You have to trust that the label is what it says it is.

Then, food preparation got abstracted into cooking as-a-service: now you don’t know if that food was cooked in butter anymore; it might be margarine, cheaper vegetable oil, or something worse. Then, the food preparation got abstracted into a menu with an “order” button”. Now you don’t even know who’s preparing it – it could be the restaurant you know, but maybe it’s a counterfeit; maybe it’s some illegal cloud kitchen, operating in a garage. The more layers you add, the more you have to just trust the label, but the more untrustworthy the label becomes. 

All internet platforms face this problem sooner or later. Amazon struggles with knockoff goods on its third party platform, while competing against upstart marketplaces like AliExpress that might have different policies. Uber and Lyft put extensive resources into verifying who’s driving on their platform, as do Airbnb with their hosts – user ratings are great when you’re dealing with good actors, but not so much for preventing abject ripoffs. Food isn’t unique. It just feels especially important, since, you know, we eat it. 

One recent trend in the food industry is large CPG companies and retailers exerting more control over their own supply chains. This makes sense; if you have a strong brand, you can’t afford the risk of letting counterfeit, sub-standard and illegally sourced food into your supply funnel too much. Some buyers are doing a better job of walking the walk than others. Costco, for instance, does a very good job of actually inspecting food, doing the hard work, and running things well. Other buyers opt for more of a theatrical approach:

Now, with all of this in mind: I think everyone agrees that the Kin Khao counterfeit restaurant situation is absolutely unacceptable and that Grubhub and Seamless cannot condone this on their platform by any means. It’s not just misleading the customer; it’s also real-time theft of the real restaurant’s brand and reputation, and sticking them with a risk (e.g. a customer gets food poisoning) that is dishonest. It’s fraud. 

But food is full of fraud! Shouldn’t we, over time, realistically expect that this abstraction layer of cooking-as-a-service will settle down to some percentage of fraudulent commerce that we just grudgingly accept? I’m not saying this is right, or okay, but I’m just saying that it seems… likely? 

Shouldn’t we expect to see some level of grift, like: restaurants that claim to cook food to a certain health standard but then don’t? Or restaurants that piggyback on a beloved restaurant that shuts down by “reopening online” (and appropriating the brand in the process)? Or virtual restaurants that create fake celebrity endorsements, drive paid acquisition with their brand until they get caught, and then shut the whole thing down and reopen the next day under a new name? 

As I said before, most of the discussion I’ve seen about food delivery apps and whether they’re good or bad for restaurants has focused on their pros and cons as aggregators. But there’s an entirely different set of questions to ask about their pros and cons as platforms: specifically, to what extent should they be held responsible for bad actors that use them? Are they more or less responsible than say, Amazon, for what gets bought and sold on their watch? These questions probably matter less in the short-term food delivery wars, but they’re more important long-term if we think about the future of food in general.

Cloud kitchens make this a much harder problem. When you’re in a physically different location from the customer, and it’s really easy to spin up a new brand and menu (while keeping the kitchen and the staff), doesn’t that incentivize a certain amount of corner-cutting, if not outright misbehaviour? To what extent can Grubhub or Seamless be reasonably held accountable for this? So long as they’re in a race for market share with one another, can they afford not to let new restaurants onto their platform? How much can they afford to spend on policing what happens on their platform?

More generally, are food delivery apps “common carriers?” Are they neutral platforms? If they’re not neutral platforms, and they’re accountable for what gets sold, then do they also become accountable for everything that passes through them, like the quality of ingredients and labour practices of restaurants they serve? And if they are neutral platforms, how much fraud and bad behaviour can we reasonably tolerate before it becomes a major problem? 

Does all of the fraud in the food industry that already existed simply become a software problem? Or does the new kind of fraud that software and the internet make possible simply become lumped in with all the food problems? Who should be policing this, the FCC, or the FDA? These are exhausting questions, because there’s no right answer yet. But the food cloud gets bigger every year, and we need to figure out whether it’s a cloud with food in it, or food that’s cloudy – and whatever that means to do next.

*Thanks to Charles for emailing me about this after I posted this on Saturday. It’s absolutely possible that this particular instance was a mixup rather than a crime, and most of the time, that’s going to be the case. But I still don’t like it! Even if this was totally a case of delivery apps making their own profiles of restaurants in order to give their customers more delivery options, that strikes me as a really disingenuous disintermediation (especially if the restaurant has no idea) – and, in the long run, feels inevitably to me like a path into outright grift.

Like this post? Get it in your inbox every week with Two Truths and a Take, my weekly newsletter enjoyed by thousands.