Brand Hijacking and Amazon's China Strategy

The true cost of cutting out the middleman

Ranjan here, and this week I'll be talking about what that cheap thing you recently bought off Amazon really means.

I've been talking to a friend who's a cofounder at a womenswear ecommerce startup about their content strategy. I searched around to see what kind of stuff is out there about them (press mentions, reviews, etc.), and stumbled upon something odd. On a top ten sex toys list, it had listed a product from their brand. They do not sell sex toys. I clicked through, and it led to an Amazon site with their company’s branding. They do not sell on Amazon. 

It turned out a China-based seller had “hijacked” their brand. This is apparently a regular thing.

A few days later, when visiting my friend's office, I found out that they had one staff member dedicated to monitoring Amazon for exactly these situations. There was a big spreadsheet where they tracked various culprits. There was a specific contact at Amazon they would call when they found shady stuff like this. They had a lawyer they billed, and a process in place to deal with this. It cost time and money and it was a never-ending game of whack-a-mole. It had become such an increasingly frequent problem over the past few years, yet they seemed fairly blasé about it. It was just business as usual.

I understand counterfeiting has always been a problem in retail, but this felt different. Amazon was their competitor. It had launched a private label brand that directly competed, undercutting them on price and shipping speed. Yet, Amazon also sold counterfeit items of theirs (well, Amazon “facilitated” it) and the startup bore the cost of cleaning up the trillion dollar company’s platform. I guess this was how ecommerce worked in 2019.


Over the past few years, I have definitely noticed things on Amazon getting cheaper, reviews becoming less reliable, product selection becoming greater, and shipping getting faster. What I had not known was, this is all part of a concerted push to rapidly increase the number of sellers, predominantly from China, on the Amazon platform.

The WSJ has been on fire covering this push:

In China over the past six years, Amazon has made its site more accessible to Chinese speakers, created special programs that address Chinese sellers’ logistical needs and sent a stream of employees to recruit suppliers.

The WSJ continued:

Starting in the mid-2000s, Amazon’s attempt to build an online retail business in China was thwarted by local competitors like Alibaba. Early this decade, it began experimenting with the new strategy, and employees “realized that global selling is much bigger” than selling in China, a former Amazon manager said.

At a Shenzhen trade fair in early 2013, no one had heard of Amazon, said Steven Chen.

Amazon employees distributed Chinese-language tutorials on opening Amazon accounts to prospective new sellers, people familiar with the company’s strategy said. Interns in Beijing phoned vendors on Chinese e-commerce sites to invite them to join Amazon.

Chinese sellers’ products often took weeks to ship across the Pacific and arrive at buyers’ addresses. So Amazon offered a logistics system, “Dragonboat,” which for a fee brought goods made in China and elsewhere to Amazon fulfillment centers in the U.S.

The piece goes on to outline how this huge push to cut out the American middleman manufacturer and go directly to the factories has led to an array of problems: a lack of adherence to safety standards, widespread manipulation of Amazon Reviews, and an overall “Wild West” feel to the platform. And that is what it has been feeling like. The article captured the sentiment perfectly: Amazon used to feel like a traditional big-box retail store. Now it was more like a flea market. 


This is where stuff started to really click. One of my earliest Margins posts was about “The Amazon Coat”. As a refresher, last winter, a nondescript coat from an unknown Chinese manufacturer somehow became the must-have item for “Upper East Side fashionista Moms”. Somehow. People marveled at how inexpensive the coat was, and always cited how good the reviews were. 

In hindsight, it seems a bit convenient that just as Amazon had gone all-in on this “straight-to-the-factory” strategy, a random coat from a Chinese furniture manufacturer became the fashion rage.

Suddenly, an entire industry of Wirecutters, NYMag Strategists, Insider Picks and a thousand others, started heavily pushing affiliate-revenue driven lists of Amazon goods. As in the example I started this piece with (a Bustle top sex toy list), the media property can push a counterfeit product, but most certainly is not liable in any legal sense, the same way Amazon is not liable for what is sold on their platform. Amazon takes their 15%, and hands off around 3% of that as affiliate revenue to whoever shared the link. Everyone gets their cut.

And if you think it sounds a bit overly conspiratorial that the “Amazon Jacket” just so happened as the company was heavily pushing this direct-to-Chinese-factory strategy, let’s remember how nuts it is that an in-demand fashionable jacked...ended up with “Amazon” in the name.. Let’s also remember, when Facebook was effectively betting its future on video, we all somehow ended up posting videos of us dumping ice water on our heads. Platforms are powerful beasts.

And the reviews! In another early Margins posts, I wrote about Amazon review inflation. That’s what made reading these WSJ investigations hit so close to home. They covered, in-depth, how this new type of seller does everything from offering Amazon gift cards in exchange for leaving reviews (something I’ve received a number of times) to the entire sport of hijacking reviews, a practice so widespread, that even Amazon’s own AmazonBasics will fall victim.

I’ve been feeling these problems of lesser product quality and harder to decipher reviews for a couple of years now. I also have been, like most consumers, getting increasingly addicted to faster and cheaper. I had a freaking 25lb kettlebell sent to my door in less than 36 hours, for a 1/3 of the price it cost at a local sporting goods store! I’m sorry. 

I just didn’t quite understand what was driving all of this. Now I do.


When China entered the WTO in 2001, it began the first phase of this era of globalization. American retailers could get rid off domestic workers and get things produced much cheaper offshore. Now, that same American retailer, the one who had laid off their employees, gets cut out, and we get to buy directly from the Chinese manufacturer. It’s a neoliberal’s dream.

For decades, we've vilified the "middleman" as an inefficiency; an unnecessary tax paid by the consumer which technology finally solved for. But, we ignored the layers of accountability and positive value that many of these conduits provided.

Think about it: If you bought a child's toy from Sears, you would assume that it didn't contain 400x the amount of lead legally allowed. But that's no longer the case:

Another musical-instrument set failing the Journal’s tests, made by a company calling itself Innocheer and listed as in China, likely contributed to a New York City child’s lead poisoning, according to city health officials. The city in May 2018 began tracking down contaminated products including the set bought on Amazon, a New York health-department spokesman says.

If you went to a store and bought a motorcycle helmet that was listed as DOT compliant, but it in fact, was not and your son died in an accident where it did not act as expected, you'd expect proper recourse, but you'd be wrong:

The coroner declared Mr. Stokes dead at the scene, a day before he and his girlfriend planned to find out their unborn baby’s gender. His mother sued Amazon, claiming the helmet was flawed. Amazon in court argued it didn’t sell the helmet but merely listed it on the seller’s behalf. It settled for $5,000 without admitting liability. It declined to comment on the case.

The examples go on and on, but you get the point. What was long decried as inefficiency, in fact, imbued some semblance of accountability. When Jeff Bezos says "your margin is my opportunity" he seems to gloss over whether there was a modicum of value being delivered within that margin.

The promise of technology was that it would remove the inefficiencies of middlemen. Yet, what's so odd is that the universal big tech platform defense has become "we're just a mindless middleman.” That they don't actually sell the product or create the content, and are just dumb pipes. 

We didn’t cut out the middlemen. We replaced them with a much more problematic and expensive version. We still pay, but in different ways. The cost of lost jobs and societal upheaval are passed on to us all. The cost of ensuring safe and compliant products is passed onto others. The cost of environmental degradation via overconsumption is passed on to everyone.

And as in the case of eliminating counterfeits, the cost is clearly passed on to their smaller competitors. My friend's company is by no means a tiny Mom & Pop shop on Main Street, Small Town USA. It's a venture-funded startup that sources from throughout Asia, clears $100mm in revenue and has over 200 employees. But it hurts my head to realize that they are the one bearing tens of thousands of dollars in costs to have to clean up Amazon's platform, and how many other comparable firms are doing the same.

For them, it's not debilitating, but for me, it's infuriating, that while having to compete with Amazon on their core business, they are also bearing the costs of Amazon’s moderation. It feels almost nonsensical when I write it out, but I guess that's how monopoly works.

DVDs and Chinatown

When I moved to New York City in 2002, whenever you’d go down to Chinatown, you would see old women selling pirated DVDs on blankets. Every now and then, you’d see them suddenly wrapping up the blanket and and running, getting caught by the cops and being arrested. Imagine if they could cite Section 230 as well.

What I’m Reading

Why the US economy isn’t as competitive or free as you think - As I already linked to a bunch of paywalled WSJ content above, I will stay premium, and link to a must-read FT piece, where Martin Wolf reviews a new book called The Great Reversal by Thomas Philippon:

First, US markets have become less competitive: concentration is high in many industries, leaders are entrenched, and their profit rates are excessive. Second, this lack of competition has hurt US consumers and workers: it has led to higher prices, lower investment and lower productivity growth. Third, and contrary to common wisdom, the main explanation is political, not technological: I have traced the decrease in competition to increasing barriers to entry and weak antitrust enforcement, sustained by heavy lobbying and campaign contributions.

Benford’s Law as a Quality of Reporting Indicator: I see some odd stuff through my work. This is one of the more intriguing things I encountered this week:

Benford’s Law states that in many kinds of numerical data, whether financial statements, population numbers, street addresses, length of rivers, etc., a predictable pattern occurs. More numbers begin with the digit 1 than begin with the digit 2, more begin with 2 than 3, and so on. Based on this pattern, the distribution of the first digit in a number is not random, it’s logarithmic, as shown below: 

Research has suggested that Benford’s Law can be used to detect anomalies in data, whether from clerical errors, random chance, or outright manipulation. When a set of numbers expected to conform to the distribution do not do so, this can be a sign that there is something wrong with the data. This simple analysis of the first digit of numbers in a data set can be used to help uncover fraud and other data problems in a number of instances, including accounting, scientific and legal cases.

It’s the weekend so let’s end with a song:

Gatekeepers gonna gatekeep

I am digging it.

Hi. This is your host, Can. Today, I’m going to tell you a story.

You know what would be the worst thing about a Zuckerberg presidency? The endless, endless memos and public speeches. Watching him not-run-for-presidency by visiting all 50 states was a traumatizing experience (though probably not as traumatizing for his hosts), but still, can you imagine him having a platform where he can constantly give these long diatribes about, I don’t know, chairs?

Anyway, I am clearly not a fan of Zuck memos (and neither is Ranjan), but reasonable people can disagree on this. Lots of people really liked his Georgetown speech, for example. Ben Thompson was impressed by the Fifth Estate framing and penned a long not-a-defense that touched on a lot of US history. One of the more full-throated (an odd phrase, but whatever) defenses came from the Singapore-based financial analyst and a generally excellent blogger Lyall Taylor. He really likes it (emphasis mine):

However, Facebook CEO Mark Zuckerberg recently gave a truly excellent speech at Georgetown University, which eloquently outlined a lot of the core principles I fundamentally believe in that underpin a lot of the analysis I wished to present. I was highly impressed, and recommend people view it. He has made my job a lot easier, as he has already summarised a lot of the foundational principles in an eloquent and concise manner, so I thought I would merely content myself to add just a few supplementary thoughts in this post. As I noted in a recent Tweet, the world is lucky to have such an enlightened moderate at the helm of such a powerful enterprise.

Anyway, my reasonable people comment still stands. Lyall’s blog is excellent, and I’ve been following it (while disagreeing with a good half) for a while. You should too.

I personally think Washington Post columnist Alexandra Petri had the best take on Zuck’s utterly shameful retconning of Facebook history, but then I read this piece by the Politico Editor-in-Chief, and it’s really brought home some of the problems I had with Zuckerberg’s view of the world, where there are no gatekeepers and journalist’s role is a historical aberration that we should do away with it:

If there was any doubt that those resentments linger, Zuckerberg laced his speech with encomiums to the fresh, clean air of direct democracy and backhanded swipes at the mildewed professional media. “People having the power to express themselves at scale is a new kind of force in the world—a Fifth Estate alongside the other power structures of society,” he declared. “People no longer have to rely on traditional gatekeepers in politics or media to make their voices heard, and that has important consequences.”

And obviously he’s not a fan:

It’s a sad development. Gaining a vehicle to challenge prevailing wisdom—whether from the government or the mainstream media—was certainly an advancement for society. The days of a few news outlets controlling the national dialogue don’t look entirely sunny, even in the rearview mirror. But the notion that spreading the news virally through Facebook pages and groups—even if scrubbed of the 3,500 different ads posted millions of times by Russian agents in 2016—provides a healthier source of information than the old-fashioned press is harder to sustain.

I am quoting extensively from it here, not because I want to fill up your inbox (although that’s part of it), but it’s a well-argued, and dare I say, balanced?

Look, you could, of course, say that EiC of a major journalistic organization would not be happy with the critical role of his industry being reduced to zilch by a bunch of hoodie-wearing 30 year-olds. But, aside from the fact that few like to apply that self-defense logic to big tech companies’ actions, it’s a simplistic view of the world. Canellos hits on an interesting point in his essay that goes undiscussed often.

When in a hole, stop Digging

The very first job I had out of college was at Digg. For those too young to remember, Digg was about “democratizing the news” before it was cool. It technically started out around the same time as Reddit. Digg, by some luck and skill, however, took off like a rocketship, created a bunch of internet celebrities, and it blew itself to pieces in one the most spectacular pieces of self-destruction that spanned over a few months. I wrote about it, and so did my friend Will Larson extensively. Reddit seems to be doing well, though.

The thing that’s less known about Digg is while the idea was to democratize the news and give power over headlines back to the people from editors, the reality was much messier.

Digg had characters, and yours truly in a questionable t-shirt during the V4 launch 9 years ago

Obviously, there was a lot of porn, and a whole lot of (not always mutually exclusively) spam. There were tons of manual and automated tooling to weed those out from ever reaching the front page. The list of items that showed up on the front was * mainly* determined by votes (there wasn’t much of a complicated algorithm), but the Digg editors still had final veto power. They had 15 minutes if I am remembering correctly before the next batch of stories would go live, so they’d visit all those 20 or so links in that 15 minutes to make sure nothing too salacious made it there.

However, the real mess wasn’t that Digg was not, in fact, some front-page made by the crowds, but rather, it was a bunch of groups fighting over it. There was definitely a considerable number of users (obviously not really a diverse crowd by any measure), but the real power generally rested with a small amount of people. There were, for example, conservative bury-brigades (digg, bury, get it?) that convened over Yahoo Groups (R.I.P.) and constituted what we now call Coordinated Inauthentic Behavior. Besides, however, the vintage alt-rights, we also had less weird people like the MrBabyman, who, with his weird but affable charm, commanded a ton of votes.

One of the many untold things that happened with Digg’s V4 launch was that our traffic and, more importantly, reputation took such a hit that we sort of had to wave the white flag to those folks. For years, Digg fought a silent battle for its front page with those influential groups; we needed them for the views, but we also did not like that they exerted as much control over the “democratic” front page as they did.

I bring all this up partly because it seems like Ranjan and I are on a reminiscing thing lately, but also, if you’ve been following anything online for the past 20 years or so like I have, none of this is really new.

Any and all online medium, any give-users-power-back website, any attempt at idealized direct democracy attempt online has deteriorated to a dynamic of couple powerful entities figuring out the rules of the game and then playing the rest of people for the mindless lemmings as they are. Millions of people acted like their votes, their likes, their shares, their diggs and buries mattered. The gatekeeper was more often just replaced by another one, except much worse, but they weren’t even aware that they were being played.

Look, this is not a defense of gatekeepers in the “we should have bottlenecks that control the flow of information” sense. Instead, I think it’s crucial that we realize the dynamics that govern these kinds of interactions online. We can either decide to be intentional about it, realize that we will always need some cooler heads to not prevail over but provide a feedback loop against the cacophony of the masses, or look the other way while the masses play the so-called neutral gatekeeper.

Sometimes I wonder if this desire to massively decentralize entire information flows is all driven by a weird obsession with mathematical purity, a desire to stretch the messy reality until it resembles the logical conclusions of any maxim. One of the oft-cited fallacies in the open-source movement is the Linus’ Law where “given enough eyeballs, all bugs are shallow.” You can see the same line of thinking in this flawed notion of The Fifth Estate, a group of free-speech loving, self-correcting mass replacing the whiny journalists of the East Coast. I mean, millions of eyes failed to notice Heartbleed for years, failing to secure a gaping hole on the entire internet basically, but sure, it will work this time!

Digg’s spiritual successor Reddit holds a good lesson here too. When the bombing at the Boston Marathon happened, a bunch of people on Reddit took it upon themselves to wade through all the available imagery to figure out who the culprits could be. It seemed like, for a moment, a celebration of how a bunch of loosely organized, individually incapable, but collectively qualified groups could uncover a deadly terrorist attack. Needless to say, it didn’t go well.

It’s one thing to celebrate more people given a voice. But maybe, throwing out the baby with the bathwater isn’t what we want to do.

😄Some Very Good News…

There aren’t many pieces that I am particularly proud of in the Margins archives. I write to think. If I had to pick one, though, it’d be Secret Liabilities of Data. It’s one of those that struck a chord with a lot of people, but most importantly, with me. I believe data is valuable. But I also believe it is a liability. For most companies, what they want is not the data, but what they can do with the data.

It’s not often you find a company that aligns so well with your vision. And even rarer to find such a company full of wonderful people you'd want to work with. So, allow me to share my joy with you all Margins readers. Starting this Monday, I'll be joining the fine folks at Very Good Security as a product manager. I know the “I’m incredibly excited” trope is overdone now, but yes, I am incredibly excited. Even Ranjan remains incredulous this is even real because it’s so Margins-relevant.

I mean look at what they say! It’s like they read our mind:

Three years ago, Very Good Security (VGS) embarked on a mission to prove a novel concept in the data security landscape: the best way to protect sensitive data is to not possess it in the first place. Since 2016, we have enabled our customers to adopt a “Zero Data” approach to de-scope their systems from interacting with sensitive data, while allowing them to maximize data utility and quickly achieve compliance certifications.

Looking forward to sharing more of what we working on over at VGS, helping secure the world’s information. If your company deals with such critical data, we can probably help. And of course, we are hiring!

SoftBank and The Dong

Illiquid Illusions and Market Manipulation

Ranjan here, and this week I'll be talking about price discovery and valuations.

A number of you have written in that you're enjoying my mid-aughts trader nostalgia. While I in no way intend this newsletter to become some kind of therapeutic, backwards-looking exercise, there's one story I have to tell you, and it helps provide context to the current travails of SoftBank.

It involves the Vietnamese Dong.



There are a lot of jobs out there that your parents probably won't understand. "Bank Trader" was certainly one of them. My parents assumed I picked stocks, and I tried to explain to them it was best understood as an industrial-level foreign exchange teller, like the ones they dealt with when they landed at the Calcutta airport.

While we took a good deal of proprietary risk: "I think the Thai Baht will weaken in value so I'll short it"), the majority of the job was facilitating transactions for clients and managing the adjacent risk: "A client needs to sell 100 million USD worth of Thai Baht. I will buy it from them, and then decide whether to hold it or sell it right away into the market.”

Emerging market derivatives were a whole other can of worms. In a currency derivatives trade, you never exchange the actual underlying asset or currency, but just sign a contract that you both agree to settle up in US Dollars at a future date, based on a what a 3rd party says the foreign currency was worth ("Today, the USD/THB is 30.00. I will bet you 1 month from now the USD/THB exchange rate is 31.00. If it's 32.00 then you owe me $1.00. If it’s 30.00, then I owe you $1.00”).

Hopefully, that didn't get too confusing. Currencies and derivatives are quite simple once you get the basic rules down in your head.


As one of the younger people on the EM trading desk, I'd get stuck with a bunch of the smaller currencies no one wanted to deal with. You'd get funny requests, like some corporate client asking you for a price in the Vanuatu Vatu or the Costa Rican Colon. A huge coffee chain would need to pay its growers, and you were effectively their FX teller. The amounts were usually minimal enough ( < $1 million) and it was more a side service than a profit-center.

In May 2008, one of the head salespeople stood up and asked for a price in the Vietnamese Dong, in the amount of $5mm USD. Our team looked around at each other, and my boss nodded at me. The currency only traded as a derivative and had almost never been requested, but the client was a big one. It was also a bit odd that it was a hedge fund, as normally these smaller currencies were requested by corporate clients who had operating interests in the country. After some brief resistance, my boss said I had to make a price.

I quoted them, and very quickly became the proud owner 5mm USD worth of virtual Dong. Again, because it was a derivative, I didn't actually have a bank account owning the currency, I only owned the risk. In currencies, even in emerging markets, a 1-2% move is still considered a pretty big deal. A 5% move is a huge deal. A $5mm position wouldn’t kill me, so I was comfortable waiting to hold this through the weekend.

It was a Friday, and I left for a friend's wedding in Ann Arbor, Michigan. At the Sunday night wedding, I received a call from our Asia office. The VND "fixing rate", which is the official rate published by their central bank, against which your books were marked, had just weakened by over 10%.

My P&L was now down $500k. I asked them if there were any buyers in the market, but there were none. Even if I wanted to get out of the trade, I couldn't. It was a long weekend so I just had to wait. I was scheduled to get back Monday night, and called to see if I should change my flight to the morning, but my boss said it was fine, as it wouldn’t make a difference (there would be no activity during the U.S. daytime anyways).

Monday night, US time, Asia called and the position was down another 15%. The trade was now down $1.5mm. For context, that was somewhere between 20-50% of my average yearly P&L. Given it was still May, it was a huge chunk of my year's gains (for further context, we'd get paid out anywhere between 2-7% of our year's profits - with that incredibly huge range determined by a number of internal political factors).


Tuesday morning, I went into work and the dressing down began. The head of our team was from Mexico, and I don't think he quite understood the hilarity of repeating the word Dong. "Did you not understand the risk of the Dong?" "If you can't handle the Dong why are you quoting it? I don't care who the client is."

I was a bit of an aloof asshole, as I had just taken the GMAT and planned to escape to an MBA within the next two years, which probably infuriated him more. I distinctly remember also having the thought, "I'll just leave and go work for the Obama campaign". Looking back, I have to laugh at the arrogance that I had an iota of a transferable skill valuable to a political campaign.

The conversation escalated to the head of trading, who was a 6'4" ex-marine. He was American, and I think understood the absurdity of having a heated conversation while saying the word Dong, because he was careful to shout "the Vietnamese Currency." But the yelling was even louder and he kept referencing the loss as “being taken out in a body bag”.

Finally, a buyer (another big bank) came into the market, and our leadership and I agreed we'd sell and clear the position at at the massive loss. A few days later, the currency started to rebound.

The weirdest part of the whole thing was that the actual Vietnamese currency never really fluctuated. Only this really esoteric derivatives settlement calculation did. I would later find out that the big bank who had bought it from me, was, purportedly, one of the two banks that the Vietnamese central bank would consult in setting that reference rate. This wouldn’t have been illegal, though. It was just part of the game.


I bring all this up, partially, because any testosterone-driven human drama involving the Dong makes for good newsletter fodder, but more so, with regards to the ideas of illiquidity and market manipulation.

I know this will sound incredibly insensitive, but when trading emerging market derivatives that was "part of the fun". If you were a big player, you could move entire markets. As in the case above, sometimes you lost, big. That poker-like gamesmanship was what gave you the rush.

It would come out years later that there were entire chat rooms (which I was luckily never invited to) that would, in fact, rig similar fixing rates. People went to jail.

It was seared into you that the only true price is the one where you can get out. The rest is just a psychological game. So this whole tech private market valuation thing has always felt a bit weird to me.


I know we're all laughing at that SoftBank pitch deck, but it's only been a few months since we all said that WeWork's valuation was $47 billion. But it never was, and WeWork's value did not drop by 75%, or whatever it was. That was all an illiquid illusion. Trading emerging markets, where there were only a few players, was the wild west, but WeWork was way worse. At least there were buyers AND sellers.

In private markets, there are only buyers, which invariably distorts the market with an upward bias. The only selling pressure comes from the cash constraints of the underlying company. WeWork was even crazier because there was only one player controlling the entire market.

And it gets much weirder. That one buyer, SoftBank, normally would be constrained by the amount of cash in its coffers. But in this case, the Vision Fund had essentially unlimited access to the balance sheet of SoftBank Corp., the large, publicly traded Japanese telecom.

That last push in January 2019 which skyrocketed the value from $20 to $47bn came, not from the Vision Fund, but from..SoftBank Corp. I think? (from TechCrunch)

As it stands, including the $2 billion that WeWork looks to receive from SoftBank imminently, SoftBank will have sunk $10 billion into the company. Perhaps it’s no wonder that the newest $2 billion is not coming from the Vision Fund but from SoftBank directly. (Son sometimes invests off SoftBank’s balance sheet directly, for expediency’s sake and, presumably, in a case such as this one, when there may be pushback from Vision Fund investors.)

(Note: how have no financial journalists broken down the crazy commingling taking place between Vision Fund and SoftBank Corp.?)

So you have a market with one buyer and no sellers. That one buyer has unlimited access to cash. And then, let's throw in that SoftBank was in the process of raising a second $100bn+ Vision Fund, and the higher the paper profits, the easier it would be to raise it. And I thought trading Dong derivatives was a shady, opaque market ripe for manipulation.


A few years ago, a friend who is a partner at an early-stage fund bought a new apartment that was not cheap. His fund had "been doing very well", but I was always confused because, while they were invested in some new unicorns, none of the investments had achieved a liquidity event (there were some smaller acquisitions, but none of the “big winners”). He explained to me that their early success on paper had allowed them to raise a much bigger second fund, and the management fees alone allowed for the apartment purchase. He noted that when they started seeing the “real exits” was when the “real money” would be made. His apartment, however, felt very real when you walked in.

Note #1: When I first learned that the head investor of the Vision Fund, Rajeev Misra, was not from the world of private capital investing, but rather from the bank trading world, I was a bit suspect. While having made some money certainly made me think I understood private-market tech investing, it didn't. Our world was pure risk management propelled by bits of financial engineering and accelerated by lots of gamesmanship. That is kind of what these crazy SoftBank rounds have felt like. (My co-host Can had linked to this great WSJ story on Misra last week).

Note #2: While that Dong debacle hit my P&L hard in mid-2008, I ended the year with the highest profits of my career. Everyone in our whole group did. That's one of the great rubs of the financial crisis. If you were at one of the last banks standing, the only ones with good credit lines, you became the only game in town. And, while the bonus percentages declined significantly, everyone still got paid something.

Note #3: This new SoftBank WeWork turnaround deck is bananas, but again, we’ve seen this madness since back in June 2010! The slide below, from the 2010 vintage, really feels like it could be the album cover for the new band I want to create after writing this piece, Illiquid Illusion.


Airbnb and the Markets

Two sided markets are bad now?

Hi. Can here. Today we talk about two-sided markets and Airbnb

Part of the allure of traveling is having to deal with mishaps. You want some stuff to go wrong, so you’ll have stories to tell about them. There’s nothing more boring than telling your friends how you flew to the Maldives and chilled on a beach for a week. It’s the “we had to sneak into a train in Thailand after losing our luggage in Malaysia” stuff that makes it interesting. Stuff going wrong is fine. But I guess there’s a limit? Airbnb has ideas:

When they arrived, they realized it wasn’t. The front door had been left unlocked, which Patterson described as “creepy,” and it was dirty and filled with furniture that looked like it had been picked up off the street. The couches were tattered, the armchairs were burned by cigarettes, and the tables were banged-up—details I confirmed through photos she’d taken at the time.

The entire thing reads like a crime novel partly because there appears to be an actual crime syndicate behind all this. Some people post some listings. Others book them and make travel arrangements. Then the hosts cancel the listings last minute, even the guests end up having to stay at a much, much shittier place. You can see the arbitrage for the hosts immediately. When the hosts reach out to Airbnb, they generally get not much in terms of attention or money, but the scamming hosts walk away with both a rent differential and possibly some cash from Airbnb too.

There are multiple ways to react to this. 

A cold, economics-as-abstraction way to look at it is at Airbnb’s scale, this kind of stuff is bound to happen. According to its website, Airbnb has more than 7 million listings, with 2 million people staying per night. At that scale, as my former boss liked to say, you are going to run into one-in-a-million events multiple times a day. Should each and every one of those incidents be a crisis?

Ugh, yes? I talked about having empathy for your users before. The good thing about technology companies is that they can deliver stuff at a tremendous scale with basically pressing some buttons. But with that ability to abstract out complicated things like humans into rows in databases, it’s essential not to lose sight of what those rows represent. You would imagine a company who is in the business of hospitality would be, like, woke about this? No?

Anyways. But there are also other ways to look at this. Somewhere around 2015 or so, we stopped calling these companies part of the “sharing economy” but instead switched to a somewhat cooler sounding “two-sided marketplaces.” It’s also more accurate. The term is mostly used to describe the relationship between credit card processors, vendors, and end-users, and it does apply pretty cleanly to model Airbnb also.

The main idea is pretty simple; you have hosts, and you have guests. You bring them together, ask them to negotiate a price for a stay, and take a little cut. There are pretty significant network effects. The more hosts you bring, the more attractive your platform becomes for guests as there are both more options and price goes down with a larger supply. And the more guests you bring, the more hosts want to get on the platform to get on that action. Pretty simple!

You have to do a bunch of things to make sure hosts and guests don’t take the transaction out of your platform (disintermediation is the technical term). Generally, you can wrap those things under Trust and Safety, which you can read as “aligning of incentives.” I mean, you should really invest in trust and safety anyway since that’s kind of a pillar of any hospitality business.

I am stylizing things a bit here, but the point is this is how it’s supposed to work on paper, but my god, it never does. The whole abstraction is what it is, a model that eschews many of the working details that make or break these businesses. The technology that goes into these businesses, having worked on them, is not *that* complicated. It’s big and gnarly, and there’s a lot of things, but you can see it all as a very fancy CRM with some real-time ERP. The operations side is where the magic happens.

Free Market All The Things

The free market types like to imagine a world where you just build a marketplace of things, and people automagically come to it and negotiate a price. Like, why don’t we do that for everything? Wouldn’t that be super efficient?

Let’s go back to Airbnb. You have guests, who need to come to your website or download your app, and browse around a bit. There is a bunch of customer acquisition things you can do here, and historically (while maybe not always ethically) Airbnb folks have been pretty good at that. I am not saying customer acquisition and conversion are hard problems, but they are problems that you can generally throw money at.

The hosts, however, are trickier. You can employ many of the same customer acquisition tricks to get people to look at your service, but getting people to list their homes is a much, much higher bar than getting people to click on. You need to earn those people’s trust, have them take decent photos, enter all the details, and, ugh, I guess to make sure they aren’t doing anything too illegal. There are a lot of steps involving a lot of humans doing a lot of human things that you can’t throw money at. I mean, you can, but it’s a whole lot more expensive than spamming Craigslist users.

Some are More Equal than Others

So, this is all to say that in these two-sided marketplaces, both sides aren’t really created equal, and they do not occupy the same mindshare in an organization. You can see this by the sheer amount of trouble companies will go to acquire one side. Uber (not very successfully) got into car leasing business, and Airbnb ended up building hotels, to keep their marketplace liquid and balanced.

“Forget Hotels” says Airbnb in 2008

Ok, look, I said you couldn’t really buy “hosts” in Airbnb’s case the same way you buy guests, but that’s also a bit of an over-simplification. One of the open secrets (not so secret to drivers) of Uber, for example, is that in some markets, more than half of the driver’s income is incentive payments, paid on top of trip fare. The riders pay some fare, and Uber tops it up. This is way, way more complicated in reality, but the point is there’s a lot of money involved that’s outside the market players negotiating a price for the exchange. And wherever there’s money, there’s fraud.

One way to think about this “incentive” is the marketplace actually pressing its thumbs on the scale by injecting some inefficiency. You generally do this at the beginnings of a market’s lifecycle, when there are not enough supply to keep the demand around, but doesn’t mean you won’t do this later either. In Uber’s case, drivers are harder to recruit, and you can occasionally “pay” them out of your own pocket to keep them around for later profits.

At Airbnb’s case, if you can read between the lines to see where Airbnb is pressing its thumbs on the scales to keep the hosts happy. In the case of a cancellation, which is inherently a messy business, Airbnb has vast incentives to err on the side of the host and pay them a tiny bit more attention (and maybe money) than they do the guests.

Markets Strike Back

This isn’t a redemption of what the hosts are doing, but rather the point is that in these kinds of markets, people on both sides are shockingly good at discovering riskless profits. There are entire fraud departments at companies like Airbnb and American Express that employ thousands of people, and people still not only find ways to exploit the known loopholes but discover and sometimes for years benefit from these hidden flows. When you move money around, you will have to deal with fraud. When you press your thumbs on the scale, you are picking winners and losers.

The larger point in all this is that Airbnb exists not because these marketplaces are efficient, but because they aren’t. If they could come to exist, without Airbnb having to build it all and then take a cut, where hosts and guests could find each other and negotiate a price, you’d not need to build a giant HQ with beautiful vertical gardens to facilitate all this.

It’s a nice office!

But the flip side of that coin is that these marketplaces are pretty fragile, and they need to maintain a decent amount of slack while maintaining reasonable margins to satisfy the investors who want those “software margins.” This is not easy, and it’s not even clear to anyone if possible. It isn’t a surefire bet.

And the larger point is that these markets, or many, aren’t some naturally occurring phenomenon. There is no universal law of markets, no subatomic dynamics that govern these interactions. All these markets, like any other, remain as creations of ourselves that should benefit everyone involved. They are designed by humans. They should also be designed for humans.

Airbnb Shooting

This wasn’t the only Airbnb-related news this week. Unfortunately, 5 people were shot and killed in an Airbnb in Orinda, California, just 30 minutes out of San Francisco. Airbnb decided to ban party houses in response. I do not want to make light of a deadly incident, and speak on a complex issue I am not qualified to speak. I am sure Airbnb felt a humane need to act. It’s not their fault this happened. However, as an immigrant to this country, it makes me uneasy that the way we handle shootings is now making sure people do not gather at parties.

Reader’s Comments

Last week, my co-author Ranjan wrote about Gawker. In that piece, he wrote Peter Thiel came clean first to New York Times about funding the lawsuit behind the scenes. Long time Margins reader Ryan Mac reached out and pointed out that he and his colleagues first reported on Thiel’s funding of the lawsuit on Forbes.

What I’m Reading

Rajeev Misra Built SoftBank’s Huge Tech Fund. Now He Has to Save It: SoftBank and its Vision Fund was on the news a lot lately. This profile of Rajeev Misra, the man behind the fund, is…wonderful. It starts crazy, and then goes to 11.

Inside Joe Biden’s 2020 Presidential Campaign: Olivia Nuzzi gives the Biden campaign a balanced look. There’s a lot here, and it’d be a wonderful piece by the anecdotes but Nuzzi’s writing is so, so good that you just want to keep reading.

I Got Access to My Secret Consumer Score. Now You Can Get Yours Too: AAAAAAAAHHHHHHHHHHHHHH!

Gawker, Trump and Checks & Balances

The death of localized, sometimes-too-aggressive media

Ranjan here. Can and I both have a number of pet theories. One of my long-running ideas is that if Gawker was still around, Trump would not have won. Yeah, it's out there, but here goes.

Being a trader in NYC during the 2000s was full of the attendant douchebaggery. Knowing every off-menu item at Nobu. Lots of bottle service. Talking about how expensive your watch was. I could go on and on, but lots of sudden money invariably contributes to a douchesuite of entitlement, arrogance and invincibility. 

But within that unshakeable confidence, there was one thing that instilled a hint of fear among younger finance guys who went out a lot in NYC: Gawker.

Being far away from the media world, none of us quite understood "what" exactly it was, but with Gawker (along with Dealbreaker, and the entire world of gossip-y, local blog sites) lurking, there was always an inkling of fear that if you did something really stupid, you'd get publicly shamed.

There were two big ones I'll never forget. The first was the Models & Bottles guy, A.J in October 2006.

All the poor kid did was let himself get taped in some stupid club promo video. He talked about how Thursday nights were for going “top down, models and bottles" (but no one would ever drive themselves to a club in NYC!). He was “an analyst at a large investment bank”. 

The thing about trading floor "viral content" is, because everyone was sitting so close to each other, digital virality had a very strong analog element to it. Someone at the end of the row would stand up and shout "did you see the models and bottles guy?" and you'd shout to them to send you the link on Bloomberg Instant Messenger. You could watch information physically spread, a slow stream of laughter turning into a flood, flowing down your row of seats.

Gawker was relentless (as seen in "Our Continuing Coverage of AJ: Day Seven"). In the video, the guy really did come off as an idiot, but man, do I feel bad for him. I would guess whoever produced that video was more to blame. I'm not sure what happened to A.J., and hopefully he ended up okay.

The other was Aleksey Veyner’s Impossible is Nothing video resume. This was a sort of coming out party for the symbiosis of the blog culture and YouTube (this one was picked up by a blog called IvyGate). Some Yale kid sent a "video resume" to UBS. As someone who's professional life is focused on getting the corporate world to be a bit more original on the content front, I should probably applaud the originality.

It’s rough. He records himself bench-pressing a self-reported 495 pounds and serving a tennis ball at a self-reported 140 mph. He said he studied in Tibet under the Dalai Lama. Just the entire thing is perfect cringe fodder. It even went " mainstream" as the NYT Dealbook interviewed him.

I felt bad for both of these guys, yet I watched and laughed. But this is the thing in the mid-2000s; all of us had a little voice in our heads saying "Don't end up on Gawker." Human Resources at a bank would actually tell you in trainings, always think about your actions in the context of ending up on the cover of the WSJ in a negative light (stipple hedcut and all). That kept you in check during the workday, but the moment you walked out the door, Gawker and gang kept you in line.

And, I think that was a good thing.


Last week, I wrote about the lack of accountability that defines the current decade. The fear that Gawker, Dealbreaker and the others put it in young finance dudes was one small check on unabated douchebaggery. Those sites did push the line of decency. There were people who got very unfairly burned. There was a lot wrong with the model. But it was something.

It also represented a vision of what the media ecosystem could look like. It’s not a coincidence that the above two things happened in October of 2006. YouTube had become a real thing. But, YouTube played the role of video hosting and embedding service (kind of like today’s Vimeo) . In a way, it gave independent media an edge. While the big news companies were still working with ugly, heavy and slow video infrastructure, YouTube gave the indies an agile way to start playing with video.


The way Gawker went down could be the most Gawker thing possible. While the business-school sounding duopolization of digital ad revenue was the nail in the coffin for most independent media, Gawker would not go down that easily.

Most of our readers are probably acquainted with the story, but for a quick summary, Hulk Hogan had sex with the wife of his shock jock friend, who goes by the name Bubba the Love Sponge. It was done at the request of Bubba the Love Sponge, who secretly videotaped it. Years later, a rival shock jock broke into Bubba the Love Sponge's home, stole the tape and leaked it to Gawker. They posted it and Hogan sued for invasion of privacy.

Of course, that was not enough. Peter Thiel had been secretly bankrolling an effort to bankrupt Gawker for years, seeking out potential lawsuits as attack channels. The Hogan sex tape was to become "the case”. The lawyer Charles Harder, who also has represented Harvey Weinstein, and now represents Donald Trump, aggressively led the case which did, in fact, bankrupt Gawker.

Still with me?

Gawker founder, Nick Denton suspected there was a more powerful force than Hogan behind the case. There was, as Peter Thiel, came out as the mastermind in an interview with the NYTimes. It was a decade-long revenge plot after Gawker had outed Thiel years before.

In classic Silicon Valley doublespeak, Thiel, who made a half-billion dollars and is still on the board of Facebook, the company that has eviscerated the concept of personal privacy, wanted to destroy Gawker for encroaching upon his privacy. The guy who literally wrote a book encouraging you to build a monopoly declared the need to end bullyish practices. 

It worked. In the lawsuit that was originally filed in July 2015, finally ended them in June 2016, neutering them during that entire period.


I've been thinking about what the death of Gawker meant for our information environment for a while, but the effective death of Deadspin this week really put it into overdrive.

One of my good friends, who I would generously call, a late-adopter, messaged me how pissed he was about this. He quipped "one by one, the rotation of websites I check every morning has died off. Where am I supposed to go now?" 

My first reaction was, "you still go to a rotation of websites?" while my second reaction was "READ NEWSLETTERS DAMMIT", but it was a stark reminder.

All of those websites are gone.

That mid-2000s consumption mode of having a rotation of major media websites, indie publications, aggregators, and social networking sites is long gone. Now you just open one of a few apps.

I had taken issue with Matt Stoller's recent NYT op-ed, “Tech Companies Are Destroying Democracy and the Free Presswhich argues that Google and Facebook's duopolization of the digital advertising market has been instrumental in killing off local journalism. There’s still lives a bit of a neoliberal in me that reacted, "no one is entitled to that market".

But I think he's right. It's the centralization of the ad market, along with the centralization of all our information distribution channels that are the genesis of so many of our problems. It's the lack of competition and numerous players. It's the lack of the small-to-mid-market. It’s having aggressive voices, close to home, calling out stupidity.

The NYTimes, WSJ, WaPo and other big players are still doing incredibly important work. But the fact that they have "New York", "Wall Street" and "Washington" in their names is why they will never be fully trusted by many. It's the reason why simpy calling them "fake news" thing is so effective. The journalists are so far removed from so much of the country that many people simply discount them. “Local News” has become some weird patchwork of ad-plastered syndication that feels more Taboola than truth-to-power.

The rise of Big Tech was nearly unchecked in the early 2010s. While Big Media was barely registering the story, the smaller players like TechCrunch were cheerleaders selling industry event tickets. There was no check. Now that the industry receives, what would still be considered kid glove coverage compared to finance, energy, government and most other big industry, the participants are so unexposed to genuine accountability that they view basic critical coverage as an attack.

That lack of close-to-home accountability is where corruption thrives. It's where misinformation blossoms. It's where all sense of personal responsibility disappears. And the destruction of a business model that supported this type of work is gone. So I guess I do agree with Matt Stoller.


Mark Zuckerberg is currently on this kick where he likes to talk about the messiness of freedom of speech, but he's missing a major part of this. Gawker was a messy thing. Aggressive, muckraking, truth-telling can often be a very messy thing.

The Models and Bottles and Nothing is Impossible guys got hit hard, but that kind of media-shaming is now everywhere. Except, unlike having at least some blogger-journalist to blame, there is no one, just an algorithm. Before, there were real, human journalists who bore at least some semblance of accountability. Certainly enough accountability that they could be sued into oblivion. That's gone.

If some random individual posted the Hogan video to Facebook or YouTube, would that be free speech?

If someone had just outed Peter Thiel via a Facebook post, would it be okay?

Zuck & Co tell us that free speech is messy. Yes, it is, but he misses the part where holding power to account can also be messy. Gawker could often go too far. We’re repeatedly told that the transgressions are of Big Tech are simply the “price of innovation”, but having a robust, regional and sometimes-too-aggressive press is the price of truth, and we’re currently not collectively paying for that. It was killed on multiple fronts.

The business model that funded small-to-mid-sized consumer-focused publications has been decimated. Instead of the douchebag class being afraid, Peter Thiel made sure the writers are afraid. Aggressive, localized journalism is gone. The blogs are gone.

That's why WeWork can get to where it got. That's why our leaders can simply lie. That's why corruption can exist at every level of our society. And that's why I think the death of Gawker is why Donald Trump won.

…I mean seriously, just look at the timeline! In July 2015, the lawsuit is filed and Gawker is neutered through the entire rise of Trump.

Note #1: In 2013, Aleksey Veyner died at the age of 29, of what I can find, still unknown causes. I'm not sure what to say about this but figure it should be mentioned.

Note #2: In 2008 I emailed Gabe Snyder, then editor of Gawker about the term douchebags, which he posted. That was one of the moments that got me obsessed with digital media and made me want to leave finance. Thanks Gabe.

Note #3: In a nod to Andrew Granato’s Abandoned Internet Margins guest post, when you try to go to the original IvyGate post on Aleksey Veyner:

Note #4: As I see newsletters like Heated and Popular Information gain traction on Substack, it’s the fist tinge of hope I’ve felt in a long time.

Loading more posts…