9.1.2021 | Dan McCarthy
Software engineering culture has always leaned toward anti-authoritarianism. Many an engineer’s childhood reading list was full of books like Snow Crash and Neuromancer about lone hackers fighting back against enormous corporations, and the mythos of the underdog startup founder is practically a local religion in Silicon Valley. This is with good reason; more than almost any other profession, software engineering can concentrate a huge amount of economic power in a small group of people fighting an incumbent.
As an engineer at a FAANG company (Facebook, Amazon, Apple, Netflix, and Google), you work with extremely smart people on challenging, highly visible problems… but you also ship things on long time cycles, your impact on the product you’re building may be negligible, nothing you’ll work on is truly yours, and your projects are likely influenced by executive power struggles outside your department. That’s setting aside all of the ethical quandaries related to privacy, security, and ownership that are inherent to those companies and grating to anyone who self-identifies as anti-authoritarian on any level.
So why do so many great engineers keep choosing to take jobs at the FAANGs? I can help explain because I spent seven years convincing engineers to join Google. There are four major explanations. I can hear you yelling about the first one from here:
These are powerful arguments to join a FAANG. However, engineers’ decision-making calculus in this area is evolving dramatically, primarily because of one technology: crypto. Some of this shift relates to the alignment between traditional software engineering values and the culture of crypto projects (open-source ethos, privacy consciousness, and cross-company collaboration), and some of it relates to crypto technologies’ practical impact on the way smaller companies and projects can make money and compensate their employees.
I’ll dive into how all of the FAANGs’ advantages are changing below, but the upshot is that a side effect of crypto technologies’ entry into the mainstream will be a significant decentralization of the engineering talent market. If you’re at a FAANG today and you’re working on making an application 0.5% faster or more scalable, the cost-benefit calculus of leaving to join a crypto startup has, and will likely continue to, become a lot more attractive.
REASON 1: The existing equity compensation model drives people away from startups. Token-based compensation gives startups an additional tool to create incentive alignment with their employees, which can meaningfully change candidates’ risk/reward calculus when they’re considering joining a small company. Crypto companies have been the first to adopt this model.
FAANG companies offer top-of-market compensation packages, and in geographies like the Bay Area where the cost of housing is astronomical, it’s totally rational to optimize for a 90th-percentile base salary and an equity stake that offers high predictability.
In contrast, the value of startup equity has historically been extraordinarily volatile: it’s either worthless or extremely lucrative, with few outcomes in between. Even in the event a startup becomes successful, it can take a decade for stock options to reach liquidity, and many moderately profitable startups choose to stay private, leaving their early employees with little to show for it asset-wise but a deep wardrobe of logo T-shirts. If you’re a senior engineer considering a startup offer against a FAANG company, you have to factor in that time and uncertainty.
Every person’s appetite for risk is different, but the upshot is that in order to be confident that a startup equity package will deliver FAANG-competitive compensation over any time period, you’ll need to believe that:
The vast majority of startups can’t reasonably meet this threshold, particularly for senior engineers. When factoring in the timeline to reach liquidity, the risk/reward proposition that stock-based startup equity compensation offers is sufficiently extreme that most people who have mortgages and families can’t stomach it.
Replacing stock-based equity with equity in the form of cryptographically generated tokens solves this problem. In a token-based vesting model, employees accrue an ownership stake in the company over time just like stock options, with the critical differences being that tokens:
This is hugely advantageous for employees. Tokens don’t guarantee a short-term payout, but they preserve employees’ optionality to get some cash out of their equity on a shorter time cycle while hanging onto the possibility of a lucrative exit down the road.
It’s worth noting here that the complexity and inefficiency of the stock equity model isn’t entirely bad for startups. Denying employees liquidity provides an advantage to founders and investors who need workers but can’t pay them market value up front for their services. Of course, employees can sometimes sell their illiquid equity on secondary markets – but that typically requires company and investor permission, which (surprise) they usually refuse to grant.
So why would startups switch to token-based equity? The short answer is that token-based equity is so much simpler than stock-based equity that over time, employee preferences will force startups to shift. When equity is governed by smart contracts, employees can be confident that they’re making informed decisions, which is not at all the case with stock option agreements. Most crypto organizations use standard, open-source vesting contracts that unlock tokens programmatically, allowing employees to understand precisely what they’re getting and when they’re getting it. This is a stark contrast to stock option models, where employees are usually given a lengthy document written by company attorneys, advised to consult an attorney of their own, and told to hope for the best.
I believe most startups will eventually tokenize their equity, but crypto startups have already started to make that leap. A high-performing 25-person crypto company can now credibly claim that the compensation packages they’re offering are breaking even against Google in terms of value and liquidity after one year instead of seven or eight because they include a token component. As more startups follow suit, the risk/reward calculus of joining a startup will look very different.
REASON 2: All our content is controlled by a handful of platform companies, which exert enormous control over what reaches eyeballs and rely on invasive tracking of user behavior to make money. Most engineers share ideological values that aren’t compatible with this model, but there hasn’t been a viable alternative for some time. Crypto can help facilitate a return to the decentralized, open web.
The web sucks now. Virtually all our public discourse goes through a few products, which are subject to abstruse Terms of Service agreements and the resultant wars over who can whine loudest about being deplatformed.
The original promise of the Internet, going back to the Usenet days, was the idea that people around the world could build independent communities focused on niche interests (minor league baseball, Command & Conquer games, kettlebell workouts) and share content without having to depend on central authorities for distribution and moderation. That idea is more or less dead. Communities based around specific interests still exist online, but at this point they mostly live within places like Facebook, which is less interested in improving your overhead press form and more interested in selling you workout equipment through a retargeted ad based on your browsing behavior.
It’s worth considering why this centralization happened:
People who want to work on big, difficult problems but also care about privacy and transparency are in a bind. As an engineer who recently left Google for a crypto startup told me:
We were doing some interesting technical work, but a lot of the hard problems came down to the scale of the data that was being aggregated or the number of people that had to coordinate to get something done, and a lot of the time was just spent learning proprietary technology or domain knowledge. The work either didn’t feel novel (migrating some project from framework 1 to framework 2) or it was advancing a mission I didn’t care about (enhancing the privacy of an advertising platform in some incremental way).
The standard startup criticism of working at Google or Facebook is that you’ll spend months working on the shading for a button. That may be true, but a lot of people will use that button – so many people, in fact, that you can virtually guarantee that it will be the most important product you’ll work on in your entire career.
Imagining a model that competes effectively with this status quo is tough. How is it possible to build a large, highly visible enterprise that’s simultaneously decentralized, transparent, and open?
The alternative offered by crypto’s on-chain governance systems is the decentralized autonomous organization, or DAO. DAOs are transparent, programmatically governed decision-making structures that coordinate large numbers of people. There are already several examples of DAOs solving hard coordination problems within finance (Compound, Synthetix, and Uniswap), but the DAO concept is extensible to many other applications – including community management.
A DAO is often the end state of a crypto startup. Crypto companies’ growth plans during their early stages are fairly traditional (raising VC money, finding product-market fit, and scaling a team). However, their long-term plans often include a transition into a fully dissolved, community-owned protocol.
Working for a DAO or protocol, or a company that plans to become one, will ultimately offer much of the same value as a large company: meaningful financial rewards on day one, equity in the project’s success, and the opportunity to work on hard problems with smart coworkers. The major difference between a DAO and a traditional private company is that the DAO’s ability to drift toward an ideological stance that its members oppose will be much more tightly constrained because the DAO is programmatically bound to decide certain things democratically.
As companies run by transparent rule sets become larger and more user-friendly, they’ll be able to offer engineers work that’s both visible and aligned with the values they believe in, with the extended possibility of creating communities that don’t have to answer messy ethical questions to scale. As a competitor to the “inject ads into a huge distribution platform for other people’s content” model emerges, big companies will have to return to their business model from the mid-to-late 2000s: making things people want.
REASON 3: Engineers want to work on groundbreaking technologies with the potential for real-world impact. Over the past decade the most cutting-edge research has taken place in secretive labs within the FAANG companies… except within crypto, where research is built on open-source principles and public experimentation.
For the purposes of this section, I’ll define impact as “the ability of one person to influence the direction of a project or technology.”
While startups spend a lot of energy selling engineers on their ability to generate impact, a lot of the most cutting-edge research still happens within the FAANGs. Engineers generally want to work on the coolest, newest thing, even in a limited capacity. Over the past decade the FAANGs have done an excellent job of bringing new research related to, for example, virtual reality and self-driving under their respective umbrellas. This keeps their brightest people working in-house, even if they’re just applying more general engineering skills to experimental products, like doing infrastructure engineering work for Google Brain. Google in particular is extremely good at identifying a technology that’s about to tip into mainstream relevance, finding the ten best people in the world in that field, and giving them all the resources they need to attack a problem, even if Google wasn’t the first mover in the space.
Conversely, research in crypto is largely done in public forums, and is often published in blog posts and open access papers. That openness creates tighter feedback loops. The foundational knowledge necessary to push the technology forward is available to anyone with an Internet connection, as opposed to being warehoused in a sealed-off building with its own cafeteria that nobody else within the company is allowed to use.
For engineers looking to optimize their career moves for impact, crypto is much more accessible than other extensions of software engineering, and it’s less likely to be locked down by large companies. The pace at which crypto is developing also means that nobody’s an expert forever, which favors newcomers – six months of work will likely put someone in the top 1% of expertise in a given sub-discipline. As crypto projects become increasingly mainstream, the population of world-class engineers in the field will grow exponentially and the pace of innovation in crypto will fly past projects that are confined within closed shops.
REASON 4: The FAANGs have historically enjoyed a built-in advantage with any candidates who need visa sponsorship to work in the US. This will continue to be a major challenge for all startups, crypto and otherwise, but in 2021, residence in the US is less important than ever before.
The immigration issue is worth an entire post on its own, but here’s a quick summary: large numbers of highly talented people want to work in America (and stay here in the long term), and it’s in everyone’s best interest to get them here. However, once they arrive, their immigration status effectively locks them to enormous companies. China and India are the two largest sources of foreign technical talent, and the vast majority of engineers from those two countries are here on H-1B visas. Given the ridiculous backlog in the visa system, H-1B holders often have to wait five years or more for a green card… and that’s if their employers began the complex and expensive application process the day they joined the company. The final phase of the green card process by itself can also drag on for a year or more, and changing jobs during that phase can cause it to restart entirely.
All this red tape creates powerful economies of scale that favor large companies with respect to visa applications. The FAANG companies have in-house lobbying arms and legal functions with expertise in immigration law, which allow them to make credible arguments to employees that staying put will guarantee their path to citizenship. Meanwhile, at a 20-person startup (crypto or otherwise), preparing an airtight visa application is often one of 30 responsibilities on an overstretched head of operations’ plate.
The distributed nature of crypto projects and the rise of remote work during the ongoing pandemic have made physically being in the US less critical to career growth. Still, great engineers want to come to the US, and they want certainty that they’ll be able to stay. As long as that’s the case – and as long as the visa process is a bureaucratic nightmare – the FAANGs will have a major recruiting advantage among immigrants.
For the past few months, the narrative surrounding crypto’s future has been expansive, to say the least. A growing number of people believe that blockchain-based technologies will rapidly replace portions of the existing financial system and many application platforms.
That said, anyone telling you what crypto is going to do in the future with absolute certainty is lying. As with the internet in the 1990s, we’re still figuring out which applications of blockchain technology will work, and there are bound to be false starts that look ridiculous in retrospect. The highest-confidence conclusion we can draw at this point is that crypto is a lever that pushes toward decentralization. The most concrete evidence for that theory is that unlike many recent technical sea changes, this one is bottom-up: despite the limited number of crypto startups in Silicon Valley and the near-total absence of crypto projects within the FAANG companies, more than 8% of Americans already own cryptocurrency.
To jump from a FAANG to a crypto startup, you don’t necessarily have to be a crypto maximalist – you just have to believe that the landscape will shift enough for 2% of the FAANG talent base to join you. As has always been the case in Silicon Valley, when the right technology arrives, a handful of smart people is more than enough.
Disclaimer: This post is for general information purposes only. It does not constitute investment advice or a recommendation or solicitation to buy or sell any investment and should not be used in the evaluation of the merits of making any investment decision. It should not be relied upon for accounting, legal or tax advice or investment recommendations. This post reflects the current opinions of the authors and is not made on behalf of Paradigm or its affiliates and does not necessarily reflect the opinions of Paradigm, its affiliates or individuals associated with Paradigm. The opinions reflected herein are subject to change without being updated.