Fabrice Grinda

Miesiąc: czerwiec 2022

Crypto-Enabled Marketplaces

By Fabrice Grinda and Matias Barbero

As we covered in FJ Labs’ Crypto Credentials, we’ve been putting a lot of thought into crypto marketplaces, especially on how traditional marketplace dynamics can change through the use of token incentives.

There are crypto-focused and crypto-enabled marketplaces, which may or may not overlap. The former is a marketplace for crypto assets, while the latter is usually a non-crypto marketplace (all offline marketplaces for instance) that uses crypto mechanics. Crypto-enabled marketplaces do not require users to buy tokens to use the platform. The crypto elements are typically abstracted from the use of the platform to maximize adoption.

Opensea, one of the largest NFT marketplaces, falls into the first camp. Users trade crypto assets, NFTs, but its underlying mechanics are not that different from, say, eBay, a centralized C2C marketplace. By comparison, Helium, is a two-sided marketplace connecting providers of wireless networks with buyers interested in using it to power different IoT devices. It deals with physical telecom equipment, not crypto assets, however, it uses token incentives making it crypto-enabled. At the intersection of this Venn diagram, we find projects such as Sushiswap or Opyn, a token exchange and defi derivatives platform, respectively, which use crypto mechanics to operate crypto-focused marketplaces.

Today crypto-focused marketplaces are more common, but over time crypto-enabled marketplaces should become the far larger segment.

In this post, we’re going to focus on crypto-enabled marketplaces and talk about how token incentives and crypto ecosystems can be very powerful tools for entrepreneurs building marketplaces. We will also highlight that using crypto mechanics should not be an end in itself and that token incentives have varying degrees of success based on the specifics of the marketplace.

Much has been written about the marketplace´s chicken and egg problem, or how to bootstrap liquidity at inception when there are barely any users on either side. There are plenty of tactics and playbooks to kick-start feedback loops: spend lots of money on marketing, build a sales team, build a SaaS tool and give it away for free, focus relentlessly on the “hard side,” subsidize through coupons, referrals, among other things. Token incentives are one of the most interesting and unexplored options in today´s toolkit.

Tokens to attract talent and have lower operating costs

In traditional marketplaces, startups must hire many richly compensated people in cash and equity to cover all of the functions of the company: product, development, sales, marketing, finance, customer care, etc.

Crypto-enabled marketplaces can behave like open-source projects, outsourcing high-impact work to the community. You can end up with much larger virtual teams and scale a lot faster while having much lower fixed costs as you use the unsalaried community to do a lot of the heavy lifting.

Given this lower cost structure, crypto-enabled marketplaces can often undercut traditional marketplaces in terms of take rate. For example, Braintrust, a crypto-enabled tech talent staffing marketplace, can charge clients an industry-low 10% fee. In fact, it actually pays their talent a negative take rate by rewarding them with tokens. Note that crypto elements are hidden form the demand side and there’s no mention of tokens or anything crypto focused: it’s all about hiring top quality talent.

For contributors, the rewards are also more immediate than if they were employees in traditional companies. Tokens can vest sooner, and the liquidity event is much faster than with traditional tech companies which are taking longer and longer to go public.

Users and contributors can not only be owners in an economic sense, but also from a voting and governance perspective. This decentralization of decision making, further aligns incentives and creates a setup of constant innovation that’s effectively managed by the community.

Tokens as a customer acquisition strategy

Think about a traditional listing marketplace that’s just getting started. Buyers will not go to the platform as there are no items on sale, and sellers are not incentivized to participate as no one is there to purchase. As a result, startups often spend a large percentage of the funds they raise  on sales and marketing to scale. Giving high token rewards early in the life of a marketplace can be a viable alternative. Note that token rewards should be highest when the overall value of the network is lowest; they come in to fill the 'chicken and egg gap’ from the get-go and should be tapered as the network effects kick in.

With token incentives, marketplaces can bootstrap one or both sides of a marketplace without appealing to heavy discounts or exorbitant marketing expenses, achieving better unit economics through a lower CAC further allowing them to undercut traditional marketplaces with lower take rates.

Instead of spending money in sales and marketing, you can provide rewards directly to the supply and demand sides of the marketplace. For instance, you can design the equivalent of giving stock options to early Uber drivers for joining the network, and for bringing other drivers and passengers.

Tokens to hyperscale

In certain cases, you can use token rewards to hyperscale the marketplace. Common wisdom in Web2 marketplaces calls for focus on “atomic networks”, very specific and niche networks as opposed to broad and all-encompassing rollouts. Andrew Chen provides great examples in his Cold Start Problem book. He mentions Uber using a hyper-narrow atomic network in the beginning to create enough density before focusing on larger regions; something like “5pm at the Caltrain Station at 5th and King Street”. Through a bottom-up approach, crypto marketplaces can be expansive and borderless while also reaching atomic networks at scale.

Helium had been around for a while before it introduced token incentives in 2020. Its main premise is to deploy hotspots in people’s homes to create a wireless network in a decentralized manner. This feat would otherwise require billions of dollars in capex from a telecom company. The graph to the left shows the exponential growth in total hotspots installed to date driven by a token incentive mechanism. Nodes went from zero at the start of 2020 to 55k+ just 18 months later.

A word of caution

Incentive mechanisms within a crypto ecosystem need to be well thought through to avoid attracting the wrong customers who are in it just for the rewards. Tokenomics can make or break a project. Web2 marketplaces also suffer from this practice when they rely too heavily on discounts to attract early customers. That said, with a robust design you can reward sticky behavior and benefit from healthier cohorts; those receiving the token incentives are enticed to let their rewards compound as they keep contributing towards an increasingly valuable network.

Also consider a potential “endowment effect” at play. Through economic and voting power, tokens could imbue users and contributors with a sense of belonging, attributing more value to the network than what it would otherwise have.

Centralized, decentralized or hybrid?

Crypto ecosystems are a powerful tool that could have varying degrees of effectiveness in different settings. Centralized marketplaces play an important role in verifying the quality of suppliers and consumers, providing customer service, improving the product, investing in alternative marketing channels, to name a few. Some (or even all) could be decentralized but that doesn’t mean there’s always a good reason to do so.

Some entrepreneurs might choose to go fully decentralized, while others might remain centralized (and could co-exist with crypto equivalents). They could also go with a hybrid model: using token incentives for certain aspects of the business while running the rest in a centralized fashion and on Web2 rails. Taking it further, we shouldn’t expect only mere transitions from Web2 to Web3 marketplaces as crypto ecosystems will unleash new, unseen models that still cannot be mapped using today’s lens.

An example of a hybrid model is our portfolio company Delphia, a mobile investment platform that uses customer data to better caliber its algorithm. In essence, Delphia is both a traditional fintech company, allowing retail investors to invest their money in an easy manner, and a conventional hedge fund, using data to power their predictive models. There is nothing decentralized or crypto in that.

However, hedge funds typically buy data from data providers (not consumers directly), and this data quality doesn’t even come close to that of companies such as Facebook or Google. So, Delphia turned to tokens to incentivize their users to contribute personal data in exchange of token rewards. There’s an additional benefit: the larger the number of contributors, the better the algorithm will get, and the superior the investment returns of Delphia users will be. In this example, Delphia chose to use a crypto ecosystem only for sourcing data.

Factors to consider

We can think of the degree of effectiveness of a crypto ecosystem, and more precisely, of token incentives along several spectra. Each spectrum can be looked at in a vacuum but, all these variables will all interact with each other which is why it’s hard to predict in which settings tokens could be more successful.

Is your marketplaces demand or supply constrained?

Most marketplaces are demand constrained. Supply is usually easier to acquire because it’s financially motivated to be on the platform. Even if there is no liquidity, if there is no cost to being on the platform, why not join? On top of that, acquiring the demand side typically involves paying the Google and Facebook “tax” of online marketing.

However, there are many exceptions. Rebag, one of our portfolio companies selling high end designer handbags, was famously supply constrained for most of its early history with 100% of the supply on the site selling out.

Whichever case you fall into, figure out the more constrained part of your marketplace and focus incentives on that side.

Are your users tech savvy?

Given the technical complexity of the crypto space, to date the most successful crypto-enabled marketplaces have worked at incentivizing tech savvy contributors. Braintrust for instance uses tokens for its technically savvy developers (their supply side), but hides all mention of crypto from its demand side clients. In this case, the supply side is more likely to know how to handle and value the token incentives than the more traditional enterprise clients on the demand side.

What’s the nature of the marketplace’s transaction / what’s being asked from the participants?

On one side of the spectrum, we might have a very easy ask, like in Helium (“hey just buy a hotspot, install it, and forget about it while you earn tokens”). On the other side, you may have something like Masterclass, where celebrities need to put on a ton of effort to create and film the courses, so even if they get tokens in return, they may not bother. All else equal, tokens should become less effective as the effort required to participate in the marketplace increases.

How homogeneous is the target supply/demand you want to attract?

You could succeed at attracting hosts for your, say, crypto-enabled Airbnb, but what type of properties are they bringing on-line? There’s a wide dispersion of types of listings they could bring but not all will be useful for your target demand, which will result in no transactions, and churn on both sides.

What’s next for crypto-enabled marketplaces?

Marketplaces have been FJ Labs’ bread and butter since inception. Even though we started investing in crypto years ago, the industry’s maturity was such that we focused our efforts on other areas such the infrastructure layer, knowing it would take require some cycles before the timing was right for greater adoption of crypto-enabled marketplaces.

However, we feel the time has finally come. There are several projects in this category already, and we are about to witness an explosion in the use of this model. We are beyond excited to see what entrepreneurs will build in this space!

The benefits of flexibility and reading the macro tea leaves in venture

As I highlighted in FJ Labs’ Investment Strategy, FJ Labs is stage agnostic, geography agnostic, and industry agnostic. In other words, we invest in any geography, in any industry, at any stage. That’s not to say there is no area of focus. We specialize in marketplace and network effect business models which represent over 70% of the investments we make. Moreover, most of our investments in terms of number of deals are at the Seed and Series A stage and in the US. In a way that is not surprising as there are more Seed deals than A deals, more A deals than B deals and so on and so forth. Also, we are a New York based fund and there are more startups created in the US than anywhere else.

That said that flexibility has been extremely beneficial. When we identify trends, we typically invest in them in all geographies. We have also been able to adjust our strategy based on our reading of market conditions. For instance, we largely stopped investing in Russia after Putin’s annexation of Crimea in 2014. We also stopped investing in Turkey after Erdogan became president.

The place our reading of market conditions has expressed itself the most and in the percentage of capital we deployed late stage, Series C onwards. While we always invest in more early-stage deals, we often wrote large conviction checks in late-stage startups. As a result, most of the capital we deployed was sometimes in the late stage. In 2015 and 2016 over 60% of the capital we deployed was in the late stage. In 2019 and 2020 it was still 35%. In general, we feel that if the company is playing in a huge category, has a high probability of a 3x from our entry price and a reasonable probability of a 10x, it’s ok for us to enter at a multi-hundred million or even multi-billion valuation.

To give you a few examples of late-stage investments, we invested large amounts in Alibaba at $4 / share or a $15 billion valuation. We started investing in Airbnb at a multi-billion valuation and continued all the way to $17 billion or so. We first invested in Farfetch at a multi-hundred million valuation.

Sometimes we invest late-stage because we did not see the deal early, but more often than not it’s because we chose to wait until the company hit real product market fit with attractive unit economics at a reasonable valuation. For instance, we saw Coupang early but passed because they were following the Groupon model. Later they pivoted to becoming one of the companies vying to be Amazon of Korea. It’s only after they established themselves as the clear leader and hit a multibillion valuation that we pulled the trigger and started investing.

Reading the macro tea leaves

In February 2021, I published my macro piece Welcome to the Everything Bubble!. In it I argued that “The warning signs of market mania are everywhere. P/E ratios are high and climbing. Bitcoin rose 300% in a year. There is a deluge of SPAC IPOs. Real estate prices are rapidly rising. … It is unclear when the bubble will burst, but there are a few ways to be ready for when it bursts. … Right now, I am building my cash reserves while still investing. I particularly like the arbitrage of selling overvalued public tech stocks (or pre-IPO companies) and investing in somewhat less overvalued early-stage tech startups. I suspect having large cash reserves will come in handy at some point in the next few years.”

At FJ Labs we took that warning to heart. We significantly decreased our investments in late-stage companies to 11% of startups and 23% of the capital deployed in 2021. Likewise in our new fund, FJ Labs III which we started deploying in July 2021 and spans 2021 and 2022, late-stage investments represent only 10% of the startups and 22% of the capital. Even then, we mostly invested in reasonably priced late-stage companies with strong balance sheets and good unit economics which were usually follow-ons in our best startups.

We also sold secondaries in many of our late-stage companies. We loved the founders and the companies but felt it wise to de-risk our position and build up our cash reserves for the crisis to come. It is worth noting that we were usually begged by the founders and other investors to sell secondaries such that the new investors could reach their target ownership without diluting the founders too much.

This is not to say we played this perfectly. We did not sell as much in secondaries as we wanted to. We also misplayed our public equity positions. Most of them fell 50% between IPO and the expiration of the lockup 6 months later. As we felt the valuation had corrected and we loved the management team and companies, we decided to hold on to most of the positions only for them to fall another 50% since then.

However, we played our hand well and are entering this crisis in a position of strength. We deployed less than 25% of the $250M we have closed across our new funds for FJ Labs III. We plan on raising $400M+, at which point our 2021/Q1 2022 investments will only make up 15% of the investable capital in the fund giving us plenty of dry powder.

The most iconic companies of the last decade were created during the Great Recession of 2008-2009: Uber, Airbnb, Whatsapp, Instagram and countless others. We expect that many of the most important companies of the coming decade will be created in the years to come. Considering the environment, the founders will build more sound companies raising the right amount of capital at reasonable valuations, taking care of their burn, and focusing on their unit economics. Because they will face less competition, they will benefit from lower customer acquisition costs and will be more likely to dominate their category.

We beyond excited to invest in this coming generation of founders and startups whom we expect will rise to challenge of addressing the fundamental issues of our time: climate change, inequality of opportunity and social injustice, and the mental and physical wellbeing crisis.

FJ Labs’ Crypto Credentials

People do not realize how active FJ Labs is in the crypto/Web3 space. We are investors in 47 companies and are investing in both equity and tokens in up-and-coming crypto projects. In this post, I want to present how we became involved in crypto. In upcoming posts, we will cover our crypto strategy and our thoughts on crypto marketplaces.

As I articulated in my 2017 post Some thoughts on cryptocurrencies, my interest in crypto came from my exposure to Argentina with its repeated crisis. Being an avid gamer with powerful GPUs, I mined BTC out of intellectual curiosity in the early 2010s. Unfortunately, I did not take any of this particularly seriously and lost 100% of the BTC I ever mined as I did not remember where I stored them. I am pretty sure I lost more BTC than I currently own. Fortunately, I bought some on BitStamp which I managed to sell before the last bubble burst in 2017.

I was fortunate to read the tea leaves well enough to get out of the liquid market in time. As I wrote at the time: “I am extremely bullish on the future of the blockchain and cryptocurrencies. However, I also believe that most of the coins in existence today will go to 0. Many of them have no fundamental reason for being. The application they support does not fundamentally require a coin. Moreover, there is a large amount of fraud and frankly ludicrous projects that have ICO’d. ICOs will not replace venture capital. Only blockchain applications are being funded by ICOs and frankly most of the companies that ICO’d would not have been funded by proper investors. What is creating the frenzy in ICOs right now is a fundamental imbalance in supply and demand. You have large crypto holders in countries like China who are looking for assets to buy, the easiest of which are other crypto currencies. That bubble will eventually burst, though I am not foolish enough to begin to pretend to know when that will happen. Bubbles tend to last longer than people “in the know” suspect. When the Internet bubble burst, hundreds of companies failed. However, they left behind the infrastructure and some of the companies that ultimately led to the Internet revolution we are still experiencing today. The current crypto bubble will also burst. It will wipe out the value of many coins and companies, but it will have funded the creation of the building blocks of future successful crypto and blockchain applications.”

At the time, we thought through what crypto marketplaces we should build or invest in, but most of the marketplaces we were presented with did not articulate well why they would benefit from decentralization and how they would use tokens to incentivize liquidity creation. In general, it felt extremely early. As a result, we ended up having an ad-hoc approach to crypto investing:

  • We invested in our friends who were building crypto companies which led us to invest early in Animoca, Figment, Shipyard/Clipper, and Zengo.
  • We invested thematically in megatrends we believed in. Having missed investing in the centralized exchange wave, we decided to invest in the best crypto onramps and offramps like Moonpay, Ramp and Wyre.
  • We invested in many companies providing infrastructure for building crypto applications.
  • We invested, tactically, in the top crypto-native funds including Multicoin and Libertus in 2017 and 2018 to get our practice off the ground and co-invest with them.

Ultimately, crypto marketplaces emerged most explosively in decentralized finance. This makes sense, as programmable finance and code-based, disintermediated financial applications were always a natural teleological conclusion to programmable money. And, in hindsight, of course the first real crypto marketplace applications to emerge were crypto native rather than replacements of existing offline marketplaces. In 2020, with the emergence of the “DeFi Summer” decentralized finance applications started growing dramatically. It turns out that DeXes like Uniswap and Sushi, lending protocols like COMP and AAVE, and even insurance protocols like Nexus Mutual are all marketplaces. They are intermediaries between providers of liquidity and those who require liquidity. The same dynamics apply as in Web2 marketplaces making our expertise highly relevant. On top of that the rise of NFTs in 2021 led to insane growth at OpenSea, another marketplace. Both trends brought our attention back to the space in a major way.

Given how early we are in the Web3 space, there is no unifying theme to our investments. We’ve been investing at the intersection of fintech and crypto, exchanges, consumerization of crytpo, NFT marketplaces, tooling and infrastructure for crypto-native teams and DAOs. Of late we have been putting a lot of thought into:

  • Bringing Web2 UX/UI expertise into Web3 which is something Braintrust has done extremely well.
  • Verticalizing OpenSea. OpenSea is the eBay of Web3. It has a complex user interface and is not particularly well suited for certain verticals. The same way eBay and Craigslist were verticalized in Web2, it’s easy to imagine the emersion of an amazing vertical gaming NFT marketplace, or perhaps a Zillow/Trulia meets Compass for metaverse properties, and many other vertical applications.
  • Helping prepare for a cross-chain world with Layer 0s and bridges.
  • Making crypto derivatives markets significantly larger than the spot markets.

All in all, to date we invested in 47 crypto companies. All that to say that crypto is now near and dear to our hearts. I will share in upcoming blog posts how crypto falls into the FJ Labs portfolio and will cover best practices for building crypto marketplaces.

If you are building something in the crypto space reach out. We would love to hear from you.

Stealing Fire: how to harness flow states for performance

The consensus has been that extreme performance requires the vaunted 10,000 hours of deliberate practice which require tremendous grit and determination. Authors Steven Kotler and Jamie Wheal take us on a globe spanning adventure meeting with tech founders, Seal Team Six, Red Bull athletes and many more in search of a short cut for peak performance.

The various groups they speak with use different approaches and techniques, but all come to the realization that such a shortcut exists: altered states of consciousness. These states are within our reach and the book covers many paths for entering them: transcendental meditation, flow dojo, psychedelics, Tantric sex, extreme sports, and others. They are all different modalities for achieving the same state.

This book is a must read for anyone interested in neuroscience, self-improvement, or altered states of consciousness. My one nitpick would be that this book is written from a very masculine perspective in Tantric terms: it’s all about achieving a state of ecstasy for the purpose of performance. As I detail in Why? flow states and altered states of consciousness are worth pursuing for their own sake.

>