Mission and Variance

In this issue:

• Mission and Variance—Some of the most successful companies are mission-focused rather than purely financial, but at least some of that is survivorship bias. Some companies fail in their mission, and others evolve into a more conventional business before achieving it.
• Transitory, After All—Looking back on the great inflation.
• The Amazon Model—Launching lots of small bets means that someone's taking lots of small losses. Sometimes that someone is the consumer.
• Contextual Ads—When you can't target based on user data, you can still make an educated guess.
• The PE and Sports Transition—Nobody buys sports teams anymore; they're too expensive.
• Credit and Insurance—When deferring interest is a great trade.

This issue of The Diff is brought to you by our sponsors, Brex.

Mission and Variance

For the most part, owning a company is just a special category of having a job. There might be a particular way you'd like to do things, but most of the decisions are, in effect, made by outside forces: founders have total control of their product until they actually try to build it or sell it, at which point a lot of what they do is determined more by their suppliers and customers than by their own agency. Which is how it should be: the world's supply of freedom is roughly fixed at any given level of productive capacity, so in the short term the more power companies have the more that comes at the expense of whoever they're working with—for you to dictate terms to someone else, they have to be deprived of the opportunity to set the terms for that transaction themselves.

There are some exceptions to this: a small number of companies credibly insist that they exist to fulfill a mission, with profit as some combination of a) a tool they use to marshall the resources for their real mission, and b) a convenient measuring stick. If your goal is to do something truly valuable for the world, you can leave the binary question of whether you're working toward that end open—it's a matter of personal judgment, and on the question of the worthiness of your work you'll be both the best-informed and most biased observer. But it's also a good bet that if you do something of value, someone will be willing to access that in exchange for some measure of universally-accepted value, like the dollar.

The way variance gets introduced to a mission-driven company is by contradiction: there are times when a normal company would course-correct, and a mission-driven one won't. This can show up in many different places:

• When Amazon was launching EC2 in 2006, the original plan was to rent servers to customers for fifteen cents an hour. At the last minute, Bezos revised this to ten cents an hour. So it was born as a negative-margin business, but fit in with the general Amazon concept of making new products no-brainer cheap and then figuring out the business model as they scale.
• Early Google launched some products that would be hard to monetize (like Google News) or that didn't seem to have a business model at all (having Google Books digitize countless public-domain works, or providing a free CAPTCHA service that also happened to help with that digitization process). But that kind of thing does make sense in light of a mission to organize the world's information.
• It's very hard to run a free speech-focused platform, both because different governments have different ideas about what constitutes free speech—depending on where you happen to be, you can get in legal trouble for saying that the Armenian genocide happened or for saying it didn't. But even beyond that, a free speech absolutist site ends up allowing some content that other sites ban, which means it becomes a popular place to share that kind of content, which gives it the self-fulfilling reputation for being all about that kind of content. While the details are unclear, Telegram founder Pavel Durov is in legal custody in France, apparently because of moderation issues on the app.

This is not the category of problem that more reactive companies have to deal with, but that reactivity also means that they're more interchangeable. For many categories, that's just fine; the world needs grocery stores and steel mills, and doesn't particularly need these companies to have some kind of unique cultural norm.^[1] So a mission adds some frequent downside variance, by forcing companies to veto profitable but non-strategic decisions and to pursue projects with limited near-term economics. But it also creates more room for extreme upside: pair a mission with a regular business, and you have a better shot at recruiting better people, and some of these projects actually evolve into good businesses on their own.

This approach works incredibly well when there are economies of scale, whether those are from network effects, experience curves, or standardization. Any time a business has nonlinear returns to scale, there's upside from pushing it just beyond what the current economics support—Intel got big by repeatedly designing chips that performed better than what the current generation of software needed, because that meant they had a lock on the next generation; Amazon's retail business has long been a mix of mature, low-growth, high-margin subcategories and emerging revenue contributors; Meta has repeatedly let unmonetized products (mobile, stories, reels) deprive them of near-term revenue so they don't lose control of some new form of online social interaction.^[2]

The mechanism by which mission-focus increases variance is a tricky one, because what it's really doing is increasing visible variance. It's often useful to think of the economy in terms of shock absorbers: when there's some change to real output, positive or negative, it doesn't show up everywhere at once. Governments and reinsurance companies tend to have a financial shock-absorption function, where when there's a big hit from a natural disaster or a recession, they eat a substantial share of the cost and recover their financial position later. In industries with a large number of suppliers and a few big integrators, like the auto business, the big companies can sometimes bear the brunt of macro swings because they don't want their suppliers to go under. (They also have the freedom to do the opposite, but that makes it harder for them to profit as much from the next upswing.) In that sense, a mission-driven business is letting the business side absorb shocks, rather than the mission side. If it's too expensive to be profitable and to make humanity a multiplanetary species, SpaceX will probably report a loss while Boeing is more likely to dial back its expectations. This makes them bigger bets, and from an investor's standpoint they'll probably have a worse expected value than backing some mercenary company whose mission is to earn a return above their cost of capital. But investors, too, can care about missions, and for a few of them, mission-focused investment means feeling good about getting richer.

Disclosure: Long AMZN, META.

A Word From Our Sponsors

Banking that takes every dollar further.

Runway is life for startups, so why do most banks chip away at it with fees, minimums, and delays? Founders deserve better — that’s why Brex built a banking solution that helps startups take every dollar further. With Brex, you get the best of checking, treasury, and FDIC protection in one account. Send and receive money fast worldwide. Earn industry-leading yield from your first dollar — while being able to access your funds anytime. And protect your cash with up to $6M in FDIC coverage through program banks. Ready to join the 1 in 3 US startups using Brex and take every dollar further?

Elsewhere

Transitory, After All

One thing that stood out from Jerome Powell's remarks on rates Friday was his point that we really did see evidence of transitory inflation after all, in the form of a decline in inflation that coincided with a strong labor market. One way to model this is to grant that there's a population that lives paycheck-to-paycheck, and treat this as, at least some of the time, a revealed preference for working up to the point where their consumption needs are met, and not much more beyond that. In that model, the economy of 2021 was going through a supply and demand shock for leisure—on the supply side, the marginal worker who didn't need to work because of pandemic-era transfers was also the marginal worker who would have been doing the restaurant and hospitality jobs that were in high demand. Meanwhile, that same person was flush enough to contribute to said demand. In that model, the process of normalizing inflation was identical to the process by which people ran down the balances in their bank accounts, and adjusted down their spending.

That actually creates a strong incentive to aim for slightly higher inflation than the long-term target. Companies dealing with a labor shortage responded partly by paying more and partly by finding ways to reduce their labor input. (A QR pointing to an online menu with in-app ordering saves waitstaff a few trips per table, which means more tables per waiter.) They don't want to drop these labor-saving changes, and they'd have a hard time unilaterally cutting everyone's wages if they did. So for the bottom quintile, real wages are temporarily higher than they'd usually be; inflation is the simplest way for them to revert to roughly where they'd be had the economy taken a different, less pandemic-influenced path. That's obviously bad for those workers, but as this newsletter has argued before, it's better to aim for maximum efficiency (and some attendant inequality) in pretax wages, and then use redistributive policies to equalize consumption after the fact. That achieves the same end goal, but involves less economic distortion than trying to keep a large swathe of jobs paying above-market rates.

The Amazon Model

Amazon's growth persists because the company is always trying to cheaply launch minor new products, most of which will fail and a handful of which will turn out to be the kind of thing that can absorb billions of dollars of capital at good returns. A result of this is that they ship lots of half-baked products with maybe a quarter-baked economics—if your product is either going to die or sell millions of units, you can afford to ship it with bad unit economics knowing that it's going to be cheaper at scale. But that also means that some products that don't take off get downgraded after the fact: they sold a version of the Echo that, for a $10 higher purchase price and a $2/month subscription, would display the users' photos on the screen—and then killed the photo feature. One of the businesses that Amazon backed for a long time and is now harvesting gains from is ads, and they tend to put them anywhere they can—what's a box if not hundreds of cubic inches of valuable ad space? What's a video if not an opportunity to run a few targeted spots? Is a product search really worse if every merchant is paying for ads? They're a tax on the user experience, but they're a subsidy on the kinds of experiments that Amazon runs.

Contextual Ads

The online ad business is built on taking an effectively infinite supply of content, eyeballs, and advertisers, and matching all three of them up in the best way possible. The two best ways to do this are either tracking lots of data about individuals in order to identify which products they are or might conceivably be in the market for, or doing this but also using the terms they searched for to precisely target ads based on exactly what they're interested in right now. An older approach, necessary in the print and linear TV era but less common today, is to target broad demographics based on interests, but this is something Reddit is pitching to advertisers as its main advantage. The same person can be a lawncare fanatic, a collector of vinyl LPs, and steeling themselves for a root canal, and targeted ads will pick up on all of this. But contextual ads will pick up on which of those traits applies to them right now, and since Reddit also gets a fair amount of traffic from search, it often doesn't know who is looking at the site, just precisely what they're interested in. In one sense, Reddit's ad model is a throwback to running ads for cat food in Cat Fancy, but in another sense, it's search with an extra step and with a banner ad instead of text.

The PE and Sports Transition

For most of the history of investment bank, banks were partnerships whose equity was provided almost exclusively by current or former employees. Starting with the IPO of DLJ in 1970, banks began raising outside capital. The cynical reason for this was that their employees wanted to cash out, and the practical reason was that the required size of banks' balance sheets was growing faster than their returns, so their only options were to raise outside money or get left behind. For all its profitability, the industry didn't generate enough capital to meet its own needs and had to look elsewhere.

More recently, this dynamic has been playing out in sports, with various US leagues getting more open to ownership by financial institutions and consortia rather than wealthy individuals. Football is next, with the NFL allowing teams to have some private equity ownership. What's especially fun about this one is that one reason for the high valuation of teams is that they're a trophy asset: the richer you get, the lower your marginal propensity to consume, in part because there just aren't that many big-ticket expenses. Sports teams, fine art, and charitable donations are the last remaining options. Team economics have gotten better, too, and since trophy asset status involves multiple expansion, returns have been better than other asset classes. It's a nice Yogi Berra-esque situation: you can't restrict NFL team ownership to wealthy individuals any more; none of them can afford the trophy-asset premium.

Credit and Insurance

The economic function of insurance is to sell the law of large numbers as a service, converting a low chance of catastrophe into an ongoing cost. The standard way to do this is to buy an insurance policy covering some specific risk from a company that specializes in underwriting that risk, but there are other mechanisms: Grenada has triggered a clause in one of their bonds allowing them to delay interest payments in the event of a hurricane ($, FT). This is an elegant solution to natural disaster risk: from a humanitarian perspective, spending scarce money on hurricane recovery is a better option than paying bondholders. But there's also a good financial reason: for a developing economy, the peak marginal return on investment comes after there's been some natural disaster that wiped out a large amount of their productive capital, whose approximate returns are already known. So another way to look at this bond is that it's a way to dynamically allocate capital to a country at a time when it has an advantage in investing it, and then to shift that capital elsewhere once that's no longer true.

Diff Jobs

Companies in the Diff network are actively looking for talent. See a sampling of current open roles below:

• Growing edtech company with successes online and offline is looking for an influencer who can reach the parents of gifted and talented kids. (Remote, full- or part-time).
• A company building ML-powered tools to accelerate developer productivity is looking for a mathematician. (Washington DC area)
• A well funded early stage startup founded by two SpaceX engineers building the software stack for hardware companies is looking for a product marketing director with experience in data, observability, or dev tools. (LA/El Segundo)
• A company disrupting the insurance market is looking for an early career data operations specialist with high agency and decent quantitative skills. (NYC/Boston)
• A tech-bio startup creating the world’s largest database of protein-molecule interactions is looking for a machine learning engineer with experience training LLMs on large scale distributed systems (Remote)

Even if you don't see an exact match for your skills and interests right now, we're happy to talk early so we can let you know if a good opportunity comes up.

If you’re at a company that's looking for talent, we should talk! Diff Jobs works with companies across fintech, hard tech, consumer software, enterprise software, and other areas—any company where finding unusually effective people is a top priority.