HR Management & Blockchain
Durga-prasad_01
EmbracingtheBlockchainandanICO-converted.docx
AirFox (A): Embracing the Blockchain and an ICO
In June 2017, Victor Santos, the 26-year-old CEO of AirFox, paused to catch his breath as he approached the Harvard Launch Lab early in the morning, just as the sun was rising. He was considering publicly releasing an announcement that his company would undertake a dramatic pivot and change its business model. If he chose to go through with it, the announcement would have profound consequences. Santos knew that a key board member would resign and other employees would likely follow, and that he would be embarking into uncharted territory for his fledging startup. Outside the building’s entrance, Santos asked himself one more time whether it was the right decision and whether the risks would be simply too great.
AirFox was an early-stage startup that sold software to wireless carriers. The company had signed three enterprise customers, was generating revenue, and had a pipeline of incoming deals. However, AirFox was on track to run out of cash in three months. Santos had already laid off employees and dramatically reduced wages for the remaining team members. After months of trying, Santos had failed to find new investors to keep the company afloat. He believed that pivoting to use an emerging technology called blockchain and executing an ambitious financing through an Initial Coin Offering would be the best path forward for AirFox. But doubts were starting to creep into his mind. Was he willing to bet the company that he was right?
Founding AirFox
Santos was born in Brazil and moved to the U.S. with his family when he was 12. He attended the University of California–Berkeley and graduated with a BS in business administration in 2013. While still a student, he cofounded Ciao, a telecom services company with operations in Brazil. The company saw modest success and grew to millions of dollars in revenue. But after graduation, Santos wanted to work for Google to learn from the best, joining as a Product Marketing Manager in 2013 while retaining a connection to Ciao. Then, in 2014, he returned to Ciao full-time as its COO and embarked on launching its U.S.-based telecom service expansion.
In 2014, four major wireless carriers (Verizon Wireless, AT&T, Sprint, and T-Mobile) dominated the
U.S. cellular network market. These leading players sold wireless plans to consumers as well as excess
network capacity wholesale to small brands. The small wireless brands were called Mobile Virtual Network Operators (MVNOs) because they operated a virtual network rented from another company rather than owning their own network equipment and towers. They often served niche populations, for whom they could design differentiated plans and pricing, including plans marketed to lower- income consumers.1
Santos planned to launch a new U.S.-based MVNO aimed at low-income customers. Subscribers would be able to lower their monthly costs by watching advertisements.
However, Ciao soon confronted serious challenges. Ciao Brazil, the MVNO’s parent company, faced financial trouble and an exodus of management talent. The new U.S. plan also had difficulty winning a large number of subscribers, and it suffered from high customer acquisition costs due to fierce competition.
Santos concluded that his team’s core competency lay in advertising software that could subsidize smartphone plans. Rather than try to run a mobile operator and market telecom services, Santos thought it would be easier for Ciao to market an app. The app would show users ads on their smartphone lock screen and reward them per ad viewed by subsidizing their cell phone bill or paying out a cash reward.
A software-based business model would be more scalable because the business could distribute the app to millions of users without asking those users to change cell phone plans or buy new phones. Santos decided to pursue this new path as a new company, unencumbered by Ciao’s financial baggage. Santos branded the new company AirFox.
Launching and Scaling AirFox
In 2015, the vision behind AirFox was to enable wireless carriers to distribute the startup’s ad- serving software to their subscribers, integrating with the carriers’ back-end systems to handle advertising, payments, and billing. The carrier would distribute the software to subscribers by either installing it on their phones pre-sale or encouraging subscribers to download the app post-sale. Once the software was installed and the ads were enabled to display on the phone’s lock screen, ad revenue would be generated to subsidize the subscribers’ plans. The business model was a revenue split between AirFox and the carrier. End users would benefit by receiving cheaper plans subsidized by the advertising. The carrier would benefit by maintaining more competitive rates without sacrificing margins.
Santos believed that selling to carriers would offer AirFox several advantages over marketing to users directly. First, AirFox would be able to scale with less investor capital, because the wireless companies could more efficiently distribute the lock screen app to end users. Second, partnering with carriers created an opportunity to sell additional software. In particular, Santos believed AirFox could build and sell more flexible application programming interfaces (APIs)a that would allow its MVNO customers to offer different data costs for different smartphone apps, thus creating more customized offerings. The ability to customize plans, pricing, and packaging would augment the MVNOs’ natural advantage of being able to pursue niche markets in the face of competition from dominant carriers. There were over a thousand MVNOs and small carriers globally.2 Santos was confident: “They all wanted revenue streams that weren’t dependent on the underlying carrier and the distribution.”
2
AirFox’s goal was not simply to make money. Santos and the employees he recruited were also motivated by AirFox’s social mission to make smartphone data more affordable for low-income users. In a 2015 survey of smartphone owners, Pew Research found that financial constraints led 23% of respondents to cancel or shut off their cell phone service. The survey further found that 37% of smartphone owners reached their maximum allowable data limit. Americans who were less educated, lower-income, or nonwhite were especially vulnerable to limited connectivity.3 The AirFox team was energized to make the internet more affordable and accessible, initially for Americans and then eventually for billions of users globally.
Santos pursued product development, customer acquisition, and investor financing at the same time. He applied and was accepted to Techstars Boston, an accelerator program that invested in and coached early-stage startups in exchange for equity. Techstars helped AirFox with its strategy, recruiting, and business development efforts, as well as exposure to potential investors. After the program ended in the summer of 2016, AirFox closed a $1.1 million seed round from respected local angels and seed venture capital firms.
The team was ecstatic to sign their first carrier contract after just three weeks of discussions and saw it as a sign of validation for AirFox’s business model. AirFox was able to integrate its software with the pilot customer in just two months. Three more carrier deals followed, as well as a pipeline of interested contacts at other carriers. By December 2016, AirFox had built its user base with the initial customers and was generating $40,000 in monthly recurring revenue, earning $0.15 to $0.80 monthly per active paid user.
The pipeline of potential deals, however, did not convert to revenue quickly over the course of 2017. Santos explained, “The sales cycle was proving to be brutal. Among the deals AirFox successfully closed, it would take 6 to 10 weeks to close the sale. And that was only the beginning. It would be a further 8 to 12 weeks for integration, followed by 4 to 8 weeks to agree on a launch plan, and then another 4 to 8 weeks to ramp up and saturate the customer’s base of subscribers.” When launching with a new carrier, AirFox would send texts and notifications to subscribers to inform them about the new program. Generally, 20% to 40% of subscribers would uptake the app, with 70% to 80% retention.
The slow pace of converting the carrier pipeline to revenue made it clear that follow-on financing would be a challenge. Santos wanted to raise several million dollars of Series A capital from institutional investors. As he met potential backers, he found that investors were not sufficiently impressed with the startup’s progress or its fundamental business model. “We were pitching to VCs [venture capitalists] and not getting anywhere,” Santos recalled. “Our end users were low-income, and our customers were small carriers with a long sales cycle.” As one team-member put it, “We were making some money, but not enough money to hit the typical Series A milestones.”
Santos believed he could attract investors for the round if AirFox could accelerate the closing of new accounts, particularly one potentially lucrative contract with a prospective carrier that seemed to be falling into place. Santos decided to raise a small bridge loan to finance the business for a few extra months, during which AirFox could close the new deals that would solidify its case for a Series A.
In January 2017, he landed an introduction to a partner at a seed-stage venture fund. Impressed with AirFox, the partner committed $500,000 as a lead investor for the bridge. This development convinced several smaller backers to commit to the financing round, including the micro-seed fund that led AirFox’s seed round and held a seat on AirFox’s board of directors. In anticipation of the new financing, which had grown as large as $1 million, Santos began hiring. He expanded the engineering team and recruited an experienced Chief Technology Officer (CTO) from the West Coast.
3
Down to the Wire
In April 2017, Santos had secured the lead investor’s signature on the bridge note documents and sent him wiring instructions; the investor had filled out the wire and returned it with the requisite signatures. However, the bank transfer did not go through. When Santos inquired as to the status of the funds, he learned that the investor was in litigation with his business partners, and a Delaware court had frozen his bank account. With the funds frozen, the investor had to back out of the financing, causing the other bridge investors to step away as well. The existing investors who had sponsored AirFox’s seed round offered emotional support but did not step up financially because they did not have deep enough pockets to carry AirFox on their own. One team member remarked wryly, “It was nice to have our investors’ support. However, it didn’t pay our engineers’ salaries.”
The day after the financing fell through, Santos called an emergency board meeting. The company had a gross burn rate of $130,000 per month, seven full-time employees, numerous contractors, and only $100,000 left in the bank. (See Exhibits 1a and 1b for AirFox financials.) To save the company, Santos proposed a plan to the board that would dramatically reduce salaries; lay off five contractors, most of whom were supporting engineering; and let the new CTO go.
After the board meeting, Santos approached the remaining team members. He asked them to work at minimum wage for two months and then at half salary for three months afterward. To make up for the lost wages, the board offered the remaining employees triple their original stock option grants. Santos also promoted two engineers to more senior roles, including elevating engineer James Seibel to CTO.
On the financing side, Santos pursued dual tracks. He scrambled to find new cash infusions to continue operations and began discussions with larger companies about an acquisition. Santos approached his old employer, Google, about an acqui-hire (i.e., an acquisition in name only, but really a wholesale hiring of the team rather than a substantial payment for the company itself). He also spoke about mergers with several other startups that served the low-income mobile market. Unfortunately, no potential acquirer or partner was interested enough to make a satisfactory offer in such a short time frame.
Finally, Santos approached his two existing MVNO customers for funding. When an initial conversation with a customer went nowhere, Santos changed his approach when talking with the second partner. “I decided to hold my cards closer to the chest,” Santos recounted, explaining how he chose to deploy a more nuanced pitch. “I told them, ‘We have a note open. This is the opportunity of a lifetime, and based on our relationship, we want you to lead with $1 million.’” The customer decided to invest $250,000, buying AirFox another six months at its reduced burn rate. Santos was relieved to receive the new commitment, but he knew it was just a temporary lifeline. To grow the company, Santos needed more cash. However, as he pitched more and more VCs, Santos found they were still not interested. AirFox had a few more months of runway, but Santos felt unsure of his long-term plan without additional capital. He reflected, “This was the moment to step back. Breathe. And figure out, ‘What the heck do we do?’”
Brainstorming a Pivot
In May 2017, Santos asked Seibel and two other employees to meet him in a conference room, take stock of the situation, and brainstorm together about what could be next for AirFox. They decided to first focus on the assets they felt they already had. Santos recalled, “We wrote out on a whiteboard the assets we had. Technology, Team, Contracts. What can we make with these?” (See Exhibit 2 for a mock- up of the whiteboard.)
Santos recalled, “We kept coming back to ‘Who is our customer?’ and ‘What problem are we trying to solve?’ We saw that we were struggling with reaching the end user to help them with their finances by selling through the carriers. We wondered: Can we remove the carriers? Maybe take more risk? What if we sold direct to consumers? This was an option, but we would need a lot more money.”
The team considered whether they could help carriers set dynamic pricing based on cell phone usage data. They considered the advertising exchanges they integrated with, and how these exchanges might be interested in acquiring rare data on AirFox’s lower-income core segments.
Over the course of the discussions, the group kept coming back to AirFox’s users. “What is their chief problem?” Santos remembers asking. “The problem they face is affordability.” This led the group to think about making small loans, called microloans, to low-income subscribers who wanted to purchase more minutes and data.
The talk of microloans prompted Seibel to share his hobbyist interest in blockchain—a new set of software protocols that created and maintained ledgers. Blockchain was an innovation with growing popularity among technologists and investors. Advocates believed it could allow transactions and contracts to occur without relying on centralized third parties, such as banks. However, it had few proven, real-world use cases at the time.
Santos wrote the word “blockchain” on the whiteboard and asked, “What could we use this for?” “Honestly, I am not sure yet,” Seibel responded. “Right now, the most popular applications are
creating payment systems and creating new currencies. Maybe we could use it to help users pay their
bills. Or keep track of their microloans. There are a lot more uses. One day, this technology will be big.”
“OK, let’s leave blockchain on the whiteboard for now,” Santos replied. “We can research it more later.”
They broke to grab lunch. Several bites into his sandwich, Santos saw a text message from Seibel on his screen. It read, “BOOM,” and contained a hyperlink to a TechCrunch article titled “Former Mozilla CEO raises $35M in under 30 seconds for his browser startup Brave.”4
The story immediately caught Santos’s attention. According to the article, Brave was building a browser that used blockchain and had successfully raised $35 million in crowdfunded capital. “For the first time, I felt something real was going on here,” Santos remembered thinking. “The former CEO of Mozilla has a serious reputation and raised a lot of money. I decided to drop everything and read all I could about blockchain and decentralized ledgers.”
5
Blockchain
Decentralized Ledgers
The term “blockchain” referred to a succession of data records (“blocks”) that were individually time-stamped and referenced each other, forming a ledger (“chain”) of data entries.5 Rather than a single copy stored in a central sever, thousands of identical copies of the ledger were stored in various computer nodesb across a broad network.6
By removing reliance on a central server, blockchains were theoretically more trustworthy than traditional centralized database technologies. There would be no chance for an entity controlling the central computer to manipulate the data, transcribe an error, or be hacked.7 Blockchain could also theoretically be more open than other database technologies. With decentralized data, no company or government would be able to make data access difficult for its own benefit.8 Further, electronic transactions could be made without relying on trust, because the public ledger contained a full history of the transactions, sealed cryptographically and immutable.
These properties of greater trustworthiness and openness through decentralization were attractive, and some blockchain champions saw the technology as a powerful business process improvement tool. To them, blockchain could help reduce costs and decrease wasted time in processes where multiple organizations needed to share data and agree on the data’s accuracy.9 Other blockchain advocates saw society-wide implications for how entire organizations were financed, built, and managed.10 Some proponents saw it as “the next internet.”11
Two professors wrote in a Harvard Business Review article about blockchain:
The parallels between blockchain and TCP/IPc are clear . . . TCP/IP unlocked new economic value by dramatically lowering the cost of connections. Similarly, blockchain could dramatically reduce the cost of transactions. It has the potential to become the system of record for all transactions. If that happens, the economy will once again undergo a radical shift, as new, blockchain-based sources of influence and control emerge.12
Building Blockchains
Computers on a blockchain network all operated according to a protocol, which was a shared set of rules that made data mutually intelligible between network participants. A central part of the blockchain protocol was to provide incentives to encourage participants to maintain the decentralized nodes and ledgers and a consensus mechanism to enforce any rules and incentives. (See Exhibit 3 for an overview of how blockchains worked.)
Blockchain protocols generally provided a core set of guarantees:
1. Past entries would be accurately maintained.
2. New entries that followed the rules would be validated before broadcasting to all of the nodes in the network.
b The term “node” referred to a point in a communications network that could communicate with other points.
c Transmission control protocol/internet protocol (TCP/IP) —a common protocol for communication between computers across networks—was one of the key innovations that enabled the internet.
3. Incentives would ensure validators act correctly.13
Bitcoin
Launched in 2009, Bitcoin was the first blockchain protocol to meet these guarantees. After the Bitcoin blockchain was created, other blockchains rapidly followed, such as Ethereum and Ripple, inspired by the Bitcoin protocol and architecture.
The Bitcoin network was designed as a payment network. The network allowed users to pay each other with a digital currency. The digital currency was called bitcoin (and written with a lowercase ‘b’ to distinguish it from uppercase ‘B’ references to the Bitcoin network or Bitcoin protocol).14 There were no physical bitcoin coins or even digital files of bitcoins. There was only a ledger of bitcoin transactions. Bitcoins could be owned by a person only if that person had an entry in the ledger in which he or she had received bitcoin in the past and not yet spent it.15
For the payment network to be trustworthy, new entries to the public ledger needed to be validated to ensure two criteria: (a) nobody spent money from an account they did not own, and (b) nobody spent more bitcoin from an account than that account had previously received. The rules of the Bitcoin protocol were instantiated in software to ensure that both criteria were met in all new entries.
The Bitcoin protocol ensured the first criterion, that nobody spent money from an account they did not own, by using public-key cryptography, a tool that functioned like a digital signature—easy for an initial party to produce and easy for outside parties to validate, but very difficult for an outside party to forge. Bitcoin users would have a private key that allowed them to “sign” data. Anyone viewing the signature could easily identify it with the signer’s digital identity, often called the signer’s public key. As long as the user was the only person in possession of his or her private key, nobody could impersonate that user.16
The Bitcoin protocol also ensured the second criterion, that nobody spent more bitcoin from an account than that account had previously received, by time-stamping every transaction. Data of several transactions would be grouped together into a “block” of transactions. Each block of data referenced the block before it in chronological order.17 Nodes on the Bitcoin network were able to check each new block against the chain of validated prior blocks to ensure nobody overspent their existing balance.18
If there were multiple chains of prior transactions that disagreed with one another and all claimed to be valid, nodes using the Bitcoin protocol adhered to a simple rule: believe the longest chain (i.e., the longest chain is the valid one). The Bitcoin protocol demanded that computers always use the latest time-stamped block on the longest chain of blocks as the true state of the ledger. In other words, the Bitcoin protocol accepted length of the chain as a proxy for validity.19
Using the longest chain as a proxy for validity had the benefit that all computers in the network could agree on which state of the ledger was valid, without a central clearinghouse and without independently validating dozens of blocks into the past to ensure that the accuracy of the chain was unbroken. However, the Bitcoin protocol needed measures to ensure that length-of-chain could serve as a dependable proxy. This was where Bitcoin’s incentive structure came in.20
To ensure that length-of-chain was a dependable proxy for validity, the Bitcoin protocol made it costly for a computer to propose new blocks. This way, a malicious party could not falsify entries to the ledger and then cheaply create enough blocks for the falsehood to become part of the longest chain. To do so, the Bitcoin network required proof-of-work to be part of any new block validation. Performing the proof-of-work involved long calculations that were expensive in both fixed-cost computer
7
hardware and variable-cost electricity. In order to propose more blocks than the honest actors, a dishonest actor would need to control a majority of the expensive computing resources in the network. In practical terms, this would make fraud prohibitively expensive and push the validators to act honestly.21,d
In this way, the Bitcoin network stored data in a decentralized ledger of transactions and used incentives to maintain the ledger. It ensured that past data was immutable. It ensured that new entries would be checked for violations of the rules and that incentives would ensure that nodes acted correctly. The model first used by Bitcoin would inspire all the blockchain projects that followed.
Blockchain beyond Bitcoin
Running the Bitcoin blockchain came with costs and inefficiencies.
First, there were direct computing and power costs required for the proof-of-work. Running a computer to validate transactions using proof-of-work was called mining. Mining heavily used electricity by using a lot of computing power. Estimates of Bitcoin’s electricity cost varied widely, but by some estimates Bitcoin miners used an amount of continuous electricity similar to that of Ireland in late 2017. There were some proposed solutions to reducing energy consumption without compro- mising security for other blockchain technologies, but Bitcoin had implemented none of them in 2017.22
Second, the Bitcoin network was slow compared with some centralized payment networks. In December 2017, Bitcoin took an average of 78 minutes to process a transaction, and processing time could spike to over 1,000 minutes on high-congestion days.23 According to some estimates, the theoretical maximum for the Bitcoin network was 3 transactions per second. In contrast, PayPal, an online payments processor, achieved 450 payments per second on Cyber Monday in 2016. Credit card network VisaNet achieved 1,667 transactions per second on average in 2016. A spokesperson for Visa claimed that, if needed, Visa’s platform had the capacity to process over 50,000 transactions per second.24
Third, private keys were not always stored securely. Hackers were sometimes able to steal private keys, resulting in massive thefts. In 2014, hackers stole the private keys for 850,000 bitcoin (worth $450 million at the time) from Mt. Gox, the leading bitcoin exchange at the time.25 In 2016, hackers stole $77 million worth of bitcoin from the popular exchange Bitfinex. The exchange assessed losses from its user base at 36% of users’ accounts.26 To avoid the possibility of theft, some Bitcoin users began using cold storage. This kept their private keys on a physical drive or hardware wallet disconnected from the internet entirely, which made rapid transactions difficult.27
Fourth, the Bitcoin network was open and participation was pseudonymous, which allowed criminal organizations to create accounts and use the currency to finance illegal activities without revealing their true identities.28 Specialized websites, such as Silk Road, began using bitcoin and similar currencies to facilitate the sale of illicit goods, such as hard drugs.29 Security analysts suspected that the military regime in North Korea was using bitcoin and other digital currencies to avoid financial sanctions.30
d To encourage validators to actually pursue the costly proof-of-work, successful validation was rewarded with a small amount of digital currency, either by a fee from the parties or by crediting the validator with newly minted currency. (Source: Satoshi Nakamoto, “Bitcoin: A Peer-to-Peer Electronic Cash System,” Bitcoin, October 31, 2008, https://bitcoin.org/bitcoin.pdf, accessed January 2018.)
8
Despite the costs and inefficiencies of bitcoin, the underlying blockchain technology held significant promise and represented a transformational opportunity that some referred to as an innovative wave as large as the internet itself. The blockchain represented a platform with proxies for trust built in, such that peer-to-peer transactions could be conducted directly, independent from any centralized authority. The role of banks, governments, and other centralized institutions could be transformed if blockchain applications became mainstream methods for transferring money and completing transactions.31
Examples of innovative blockchain applications began to emerge around the world, taking the core innovation of Bitcoin and expanding it to create new distributed applications. For example, the Republic of Georgia was collaborating with a blockchain company to build a secure land registry on a decentralized ledger, making it tamper-proof regardless of the party in power.32 Finland began a project to implement a blockchain-based digital identification system to help securely track refugees, independent of their paperwork or passports.33 The United Nations kicked off a project in partnership with Microsoft and Accenture to build an identity system for the 1 billion refugees who did not have physical proof of citizenship or identity. “Without an identity you can’t access education, financial services, healthcare, you name it. You are disenfranchised and marginalized from society,” David Treat, a Managing Director in Accenture’s financial services practice, said in an interview. “Having a digital identity is a basic human right.”34 In 2017, the technology giant IBM deployed over 1,500 professionals working on over 400 blockchain projects.35 IBM sought to use blockchain to improve the accuracy of supply chains, financial services, and cybersecurity.36 Several large banks began developing a private blockchain to conduct more efficient financial transactions among themselves that did not require expensive and lengthy processes to validate separate ledgers of transactions.37 Several entrepreneurs built alternate versions of the Bitcoin protocol that sought to increase its scalability for everyday transactions38 or anonymity for hidden transactions.39
One of the most influential blockchain projects was Ethereum, initially conceived and proposed by Vitalik Buterin, a 19-year-old programmer in Canada, in late 2013 and launched in 2015.40 Ethereum enhanced the Bitcoin protocol by creating a new protocol that created a virtual machine that enabled software code to be implemented on the Ethereum blockchain. This innovation allowed developers to build a wide range of new applications on top of the Ethereum blockchain and internalize its benefits, without building their own blockchain from scratch.41 Developers used Ethereum and its richer computational platform to construct smart contracts, which one prominent blockchain researcher described as “little programs that execute ‘if this happens then do that,’ run and verified by many computers to ensure trustworthiness. If blockchains give us distributed trustworthy storage, then smart contracts give us distributed trustworthy calculations.” 42 Ethereum and other similarly constructed, programmable blockchain-based platforms such as NEO (based in China and launched in 2014) heralded a new era in the blockchain ecosystem. A host of new developer tools made working with blockchains easier, and a host of new investors made building blockchain companies much more lucrative.
Over the course of 2016 and 2017, entrepreneurs launched hundreds of new blockchain projects, many built using Ethereum’s platform. Some startups sought to re-create large existing online marketplaces with reduced fees by removing the expensive middleman at the center. OpenBazaar worked to create a decentralized e-commerce platform to compete with retail giants such as Amazon and eBay. Arcade City worked to create a decentralized ride-sharing platform to compete with giants such as Uber and Lyft.43 Other startups used blockchain to create new marketplaces for services that had previously been vertically integrated. For example, startups FileCoin and Sia planned to use blockchain to decentralize digital file storage away from centralized cloud computing companies like
9
Google and Amazon Web Services.44 Some companies focused on building out further infrastructure and scalability capabilities that other blockchain projects could use in the future.45
Just as the Bitcoin protocol used bitcoin as a digital currency that incentivized contributions to the network, each unique blockchain had its own currency—also called a “coin,” “token,” or “cryptocurrency.”e The coins incentivized behavior such as node maintenance or transaction validation and usually corresponded to some scarce resource (such as validated transactions for Bitcoin or validated digital storage for FileCoin).46 Blockchain projects began to sell coins in order to fund their operations. By mid-2017, the first offering of coins for sale to the public by a blockchain project was commonly called an initial coin offering (ICO).47
The sale of digital tokens did not simply attract the interest of potential users; it also attracted the attention of speculators and professional investors. ICOs were similar to the presale of a product (such as a book), in that they gave buyers the right to consume the product at a later date. However, ICOs were also similar to the initial sale of stock in that tokens were generally expected to increase in value if the blockchain project proved popular in the future. The tokens were generally liquid, traded on global exchanges that rapidly formed—such as Binance, Bitfinex, GDAX, and Poloniex—and could be traded among parties and exchanged for other cryptocurrencies or traditional fiat currency, such as dollars and yen. Investors became interested in buying tokens of projects that they thought would succeed in the long term, or that they thought would increase in value in the short term and could be quickly flipped to other buyers.48
ICO: Alluring Yet Risky?
Entrepreneurs were drawn to building blockchain products for the merits of the technology, the source of funding, or both. In choosing whether to use blockchain, they faced several considerations for and against.
Blockchain advantages Building a blockchain-based product had several advantages. First, there were the technical attributes of the blockchain technology itself. Blockchains represented an innovation in data storage and execution, guaranteeing data accuracy, transparency, trust, and openness without relying on any central individual or institution. These attributes could theoretically allow small startups to bootstrap trust in their service without a recognizable brand.
Second, selling tokens in an ICO offered an attractive source of financing. Selling digital assets allowed a startup to raise capital without diluting equity or taking on debt, but instead just limiting the amount of coins the startup could sell in the future. Tokens were a liquid investment vehicle for buyers and could be converted into fiat currency quickly.49
Third, founders could use tokens to subsidize early adopters. Early adopters of new network technologies often had to bear high implementation costs and to wait for later adopters to make the network effects of the technology more valuable.50 For example, early adopters of the telephone had very few users to communicate with. Blockchain could provide early network users with tokens in reward for their usage, giving adopters a financial incentive to spread a new technology even before the majority of the technology’s eventual utility could be realized.51 One blockchain investor, Nick Tomaino from 1confirmation, described this as a “token network effect.” He wrote, “A decentralized global ledger combined with a token that’s required to use the ledger aligns incentives and creates more intense product engagement than ever seen before. The product becomes more useful as more
10
users join, and demand for the token increases as a result, which increases the value of the ownership stake for users.”52 (See Exhibit 4 for an illustration.)
Blockchain risks Entrepreneurs also had to face several serious costs and risks when building a blockchain-based product. First, a blockchain was slower and required more energy compared with traditional data storage methods. At a 2016 conference, Ethereum founder Buterin told the crowd that the whole worldwide Ethereum network ran with the speed of a “smartphone from 1999.”53 Unless founders really needed the positive benefits of blockchain, this efficiency drag would go to waste as compared with using a high-performing centralized database like MongoDB or Oracle.
Second, launching an ICO could place restrictions on the entrepreneur’s business model. For many projects, the value of blockchain was to remove the middleman and thus minimize fees for a particular type of transaction or service. The ethos of the cryptocurrency community was centered around open- source principles and the creation of public goods—asking participants to contribute based on modest incentives and the promise of being a part of an exciting new community. Thus, many blockchain startups could not look to future fees for profit. These blockchain projects, including Ethereum, simply awarded tokens to early founders, employees, and investors as their means of monetization. To profit, early contributors depended on the cryptocurrency gaining value as the project gained traction. If the cryptocurrency languished or dropped sharply, interest in the project could drop as well.
Third, once launched, a blockchain protocol was difficult to modify. This meant the founders would have limited agility to iterate on the offering over time. Early mistakes in blockchain design, such as limited scalability or hyper-deflation, could have a devastating impact for the community and be difficult to fix. Having the flexibility to run experiments was a core principle in the search for product- market fit and a powerful community of users may reduce degrees of freedom. Similarly, selling tokens often required handing over control of intellectual property (IP) to the community of users. Since the protocol of the blockchain was publicly available, other developers might even decide to make a second version of the software (called a “fork”)—perhaps with some small tweaks—and lure away the community of users. A fork might leave the original token sellers with a diminished husk of a network and a devalued currency. Further, prospective acquirers were unlikely to be interested in purchasing a company where the IP was publicly available in its entirety in an open-source format and where the community was not under the control of the company. The downside of a decentralized community was that the traditional value-capture methods of building a community (e.g., through licensing additional software as open-source leaders like Red Hat and MongoDB did) would not necessarily apply.
Fourth, executing an ICO consumed significantly more team energy than fundraising from VCs did. In a traditional venture round, there were only a handful of investor contacts for key communications, and they mostly wanted to speak with the CEO. When selling tokens to the public, thousands of potential investors from around the world expected to be able to ask questions of the CEO, engineers, and community managers (a newly critical role) and demanded a high degree of transparency. Questions—including deep technical questions—coming from many time zones could keep the team working long hours for months.54
Fifth, blockchain entrepreneurs were vulnerable to theft and fraud. Although blockchains prevented theft from a centralized system, any cryptocurrency system secured by private keys was vulnerable to the theft of those keys. Moreover, programming mistakes written into a new blockchain application might allow hackers to steal the entire value of a company. In 2016, blockchain entre- preneurs created a smart contract called “The DAO” built on Ethereum. The DAO team was very successful in fundraising and attracted over $100 million in investment in less than 30 days. However,
11
the engineers had written some mistakes into their code, which allowed an attacker to steal the majority of the contributed funds.55 Developers did not even need to make a mistake themselves to suffer from these attacks. The value of ether (Ethereum’s cryptocurrency) temporarily dropped by 30% as a result of the DAO attack.56 This drastically reduced the wealth of Ethereum founders and investors even if they had no relationship to the attacked smart contract. Additionally, there were many cryptocurrency- based scams, and potential ICO buyers were worried about losing their investments to criminals. Over the course of 2017, an anonymously run cryptocurrency called BitConnect rose to over $2 billion57 in market cap by promising investors annual returns in excess of 100% for loans. Many observers considered it a Ponzi scheme.58 In January 2018, BitConnect shut down and the market cap collapsed by over 90%.59
Sixth, raising money through an ICO could reduce the willingness of VCs to participate in future rounds. Some venture capitalists believed that offering coins would make future liquidity of equity investments via acquisition difficult. “In many startup exits, the team is acqui-hired and the product shut down,” explained one early-stage VC. “As an equity investor, I would be doubtful that a company selling tokens could be acqui-hired, because the buyer might have to maintain the protocol. I wouldn’t want to buy shares with that path to liquidity cut off.”60 This potential conflict of incentives between VCs and management could be aligned through innovative deal structures, but it required careful thought and represented new territory for both founders and investors.61
Finally, the long-term regulatory future of cryptocurrency financing was unclear. Global regu- lations were in flux. The U.S. Securities and Exchange Commission (SEC) explained that while some ICOs were truly the presale of a product, many ICOs would be considered sales of securities and hence an activity otherwise heavily regulated by the SEC.62 In September 2017, the SEC created a new cyber unit with part of its mandate to “target misconduct,” including “violations involving distributed ledger technology and initial coin offerings.”63 In 2017, regulators in China and South Korea both looked to ban ICOs. The Chinese government called them a “form of unapproved illegal public financing behavior.”64 (See Exhibit 5 for the blockchain project ecosystem, Exhibit 6 for the ICO launch process, and Exhibit 7 for a list of ICO regulations in July 2017.)
Despite the risks, entrepreneurs and investors gravitated toward blockchain projects in 2016, with
$228 million raised in ICOs. In June 2017 alone, ICO funding totaled $550 million and surpassed the nearly $300 million raised by all angel and early-stage VC funding in the same period.65 (See Exhibit 8 for funding levels.)
AirFox Considers the Pivot
A New Business Model
After learning more about blockchain, Victor Santos and James Seibel were convinced that the technology could help them bypass the carriers and sell a new affordability solution in a direct-to- consumer business model. They explained their vision for a new solution as follows:
AirFox’s solution harnesses the decentralized power of the Ethereum blockchain and advertising to give users AIR, a new token redeemable for mobile data, and eventually physical and digital goods. Users can then establish credit history through their opted-in device data, advertising behavior and browsing history, enabling AirFox to fund micro- loans via the AirTokens to any person with a smart phone.66
12
They also wrote a three-phase product roadmap:
1) Phase 1 (~Q1 2018)—Launch B2C Products: Airfox App and Browser
In the first phase, we will enable users to earn free AIR via advertising for mobile recharges across 500 wireless carriers, leveraging our partnership and integration with TransferTo. Users can earn AIR by using the Android AirFox Browser or an app they can download called the Recharge App and viewing sponsored advertising. This initial offering will generate early adoption and increase demand for Airtokens. We will target Brazil as our first market.
2) Phase 2 (~Q3 2018)—Create value and financial instruments for Airtoken users
Once users start earning, buying and redeeming Airtokens, the second phase is enabling qualified active users to receive micro-loans without needing traditional credit scores. The loans will be in AIR and collateralized by their prepaid account, using our own internal AirFox user scoring to underwrite the loans. We will also expand the use of AIR for items outside of mobile recharges. Users will be able to redeem micro-loans of AIR for physical and digital products across our ecosystem as well as conduct peer to peer money transfers with very low fees. Lenders will have a dashboard to track loan performance and access to bundled loan products with attractive returns. We may have to modify our implementation to satisfy legal and regulatory requirements on a regional basis.
3) Phase 3 (~Q1 2019)—Launch Partnerships to grow the Airtoken network
The third phase is extending the use of AIR and our advertising / micro-loan blockchain system outside of our own applications and across a direct network of publishers and advertisers. Publishers and advertisers can reward their users with AIR for engaging in their advertising and use our micro-loans system to make in-app purchases. The platform would be an external library via an API and SDKf that publishers can integrate into their applications or websites, allowing for general use.67
The target customers of AirFox’s new solution would be prepaid mobile consumers who were unbanked or underbanked, mostly in emerging markets. The AirFox team wrote out several example customer use cases, their problems, and how AirFox intended to solve those problems. (See Exhibit 9 for AirFox’s discussion of the blockchain target market, Exhibit 10 for AirFox’s discussion of target customers, Exhibit 11 for AirFox’s planned market flow, and Exhibit 12 for AirFox’s AIR apps.)
Role of Blockchain
AirFox believed blockchain would enable the new business model in several ways.
First, the transparency property of blockchain would enable users to trust AirFox at low cost. Seibel explained, “I am not sure we could do the micro-lending without the blockchain. Without blockchain, we would have to host backups and audits. We would need to make sure everyone trusts us. We would need to be taking in the money. We would be responsible for knowing the interest, and transferring it
f A software development kit (SDK) was a set of tools that developers could use to create new programs. An SDK might include one or more APIs.
13
out. We would likely have to sign partnerships with major banks. With blockchain, the money just goes in and out. We don’t need to guarantee the money is safe. There is no liability. It all maintains itself.”
Second, the permission-less property of blockchain networks would allow other companies to accept and use AIR. “Eventually, we open AIR and our platform up to more companies, so it can be valuable without us,” explained Seibel. This created the potential for AIR to be broadly adopted by other applications and platforms, thereby increasing the utility and value of the currency.
Third, the community of cryptocurrency investors was a potential source of both financing and early users. Santos explained his reasoning at the time: “We thought that crypto investors would be excited about the product,” explained Santos. “We thought the investors would want to become lenders in the platform. They would really like to loan at 30% to an individual Brazilian borrower peer- to-peer. That’s a good rate for both parties.”
AirFox was not alone in realizing the potential of the blockchain to power lending applications. Many other startups had launched initiatives in this area, including SALT (a platform to list blockchain assets as collateral for loans), Jibrel Network (a platform to convert fiat currencies and other financial instruments into cryptocurrencies for lending), and ETHLend (a platform for borrowers and lenders to transact using Ethereum-based smart contracts). Many microlending startups were investigating using the blockchain’s transparency and lack of middlemen to build more efficient capital allocation systems. Moeda had created a token-based lending system to provide loans for development projects in rural areas. Sweetbridge was a new blockchain platform over which users could lend money to themselves, and BanQu was creating a peer-to-peer lending platform where lenders could securely provide data points such as their financial history and land records to demonstrate their creditworthiness.
Reactions to the Plan
Not every team member was convinced that it was time for AirFox to change business models. One senior team member said, “We were generating revenue at that point. We had a strong pipeline. It’s not like there was zero indication our existing path would work. I thought we should stay the course. If we changed paths, we would be abandoning our customers that we had worked so hard to win.”
Another key team member told Santos:
We would be moving away from our core competencies. We would be moving away from telecom, SaaS, and advertising. We would have to be experts at public relations, investor relations, cybersecurity, and catering to a fickle consumer market. Business-to- business models are in some ways easier. You provide some value and the deal will get done. It may take some time, but it will get done. And, if we did succeed with blockchain, we would be a very different company than we had set out to start.
When Santos brought his ICO plan to the board, the response was negative. An investor board member saw the ICO as very risky from both an operational and a legal perspective.
On the operational side, the ICO would distract the company from pursuing traditional financing. It would likely eliminate the potential for an acquisition that could preserve some of the equity’s value and perhaps provide some modest upside.
14
On the legal side, the regulation of ICOs remained entirely unclear. Action from financial regulators in the U.S. could shut down the company and even expose management and the directors to substantial financial liability. According to Seibel, the legal guidance at the time was, “our lawyer hears a rumor from other lawyers about what the SEC is thinking and they share the rumor with us.” In fact, AirFox’s law firm had never been through an ICO and appeared ill-prepared to provide the company with useful counsel. To the board member, the speculative return simply did not justify the risks.
The board member told Santos that if he chose to pursue an ICO, she would leave the board to remove herself from a decision with which she disagreed and that was fraught with liability.
From ICUg to ICO
After pausing at the door of the Harvard Launch Lab, Santos turned and took a walk along Western Avenue toward Harvard Square, squinting into the early summer sun.
He had initially felt confident in the new business model, but his conversations with critical members of his team and the board left him with doubts. Their different critiques of the blockchain strategy were not unfounded. Moreover, Santos was concerned with whether his team had the capacity and skill to push forward with the pivot and complete the process. It would take several months before any signs of progress emerged. Losing even a single team member could mean failure, and alternative job offers would be very tempting to engineers earning minimum wage. Pivoting to blockchain would be a bet on an untested underlying technology, a brand-new financing mechanism with unknown legal ramifications, and his own leadership abilities. Santos wondered if this was the right gamble to make.
g ICU stands for “intensive care unit.”
15
Exhibit 1a AirFox Monthly Expenses, June 2016 to December 2016 ($ thousands)
|
|
Jun-16 |
Jul-16 |
Aug-16 |
Sep-16 |
Oct-16 |
Nov-16 |
Dec-16 |
|
Payroll (incl. tax) |
10 |
19 |
22 |
57 |
19 |
39 |
57 |
|
Software Contractors |
16 |
17 |
33 |
48 |
63 |
24 |
67 |
|
Legal & Professional Fees |
0 |
76 |
0 |
11 |
16 |
13 |
7 |
|
Advertising |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
Marketing |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
Total |
26 |
112 |
55 |
117 |
98 |
77 |
132 |
Source: Correspondence with Victor Santos, January 2018. Note: Differences due to rounding.
Exhibit 1b AirFox Monthly Expenses, January 2017 to June 2017 ($ thousands)
|
|
Jan-17 |
Feb-17 |
Mar-17 |
Apr-17 |
May-17 |
Jun-17 |
|
Payroll (incl. tax) |
21 |
46 |
72 |
30 |
52 |
25 |
|
Software Contractors |
88 |
41 |
64 |
94 |
34 |
26 |
|
Legal & Professional Fees |
5 |
2 |
0 |
7 |
0 |
0 |
|
Advertising |
0 |
0 |
0 |
0 |
0 |
0 |
|
Marketing |
0 |
0 |
0 |
0 |
0 |
0 |
|
Total |
114 |
89 |
136 |
131 |
86 |
51 |
Source: Correspondence with Victor Santos, January 2018. Note: Differences due to rounding.
16
Exhibit 2 AirFox Pivot Whiteboard, Mock-Up from Victor Santos’s Memory
Exhibit 3 How Blockchain Works
$
The transaction was represented online as a “block.”
The block was broadcasted to every party in the network.
Those in the network gave approval to the transaction.
5
The block would then be added to the chain, which formed an immutable record of transactions.
The money was then moved from A to B.
6
Source: David Yoffie and Anthony Woo, “Note on Blockchain and Bitcoin, 2017,” HBS No. 718-433 (Boston: Harvard Business School Publishing, 2017), http://hbsp.harvard.edu, accessed December 2017. Re-created by Yoffie and Woo based on Scott Manuel and Sarah Andrews, “Blockchain technology: Is 2016 the year of the blockchain?,” Thomson Reuters, January 16, 2016, https://blogs.thomsonreuters.com/answerson/blockchain-technology/, and Milton Lim, “Bitcoins, Banking and the Blockchain,” Actuaries Digital, April 24, 2015, https://www.actuaries.digital/2015/ 04/24/bitcoins-banking-and-the-blockchain/; both accessed October 2017.
18
Exhibit 4 Traditional and Token Network Effects
Source: Casewriter, based on Chris Dixon, “Crypto Tokens: A Breakthrough in Open Network Design,” Medium, June 1, 2017, https://medium.com/@cdixon/crypto-tokens-a-breakthrough-in-open-network-design-e600975be2ef,
19
Exhibit 5 Blockchain Project Ecosystem
Source: “Frost & Sullivan and Outlier Ventures Identify the 2017 Global Blockchain Startup Map,” press release, Frost & Sullivan, March 27, 2017, https://ww2.frost.com/news/press-releases/frost-sullivan-identifies-2017-global- blockchain-startup-map, accessed July 2018.
20
Source: Ramana Nanda, Robert White, and Alexey Tuzikov, “Initial Coin Offerings,” HBS No. 818-067 (Boston: Harvard Business School Publishing, 2017), http://hbsp.harvard.edu, accessed December 2017. Created by Nanda, White, and Tuzikov based on Chris McCann, “Guide to launching an Initial Coin Offering (ICO),” Medium, October 11, 2017, https://medium.com/@mccannatron/guide-to-launching-an-initial-coin-offering-ico-94587af2c8d5, accessed
October 2017.
21
Exhibit 7 Select Global ICO Regulations as of July 2017
|
Country |
Regulations |
|
United States |
The Securities and Exchange Commission (SEC) had ruled that the tokens of DAO were securities, and recommended the Howey test and legal counsel for new ICOs. (The Howey test, which was often utilized to determine whether an offering was a security, would be used to structure a token so as to ensure compliance with securities and regulations.) |
|
United Kingdom |
The United Kingdom had established a regulatory sandbox to conduct pilot studies of new financial projects. The Financial Conduct Authority (FCA) was adopting a wait-and- see approach to distributed ledger technology. |
|
China |
The People’s Bank of China, China’s central bank, had announced a blanket ban on all ICO funding. A notice issued by the China’s National Internet Finance Association, a working committee regulating the country’s Internet finance sector, stated that new projects raising cash through cryptocurrencies were prohibited, and authorities would clamp down on related fraudulent practices. |
|
Russia |
The Bank of Russia, the nation’s central bank, had warned about the danger of ICOs with a statement highlighting the “high risks” of exchanging cryptocurrencies. The central bank was also prohibiting any cryptocurrency trading on any official exchanges in the country. |
|
Switzerland |
According to the Swiss Financial Market Supervisory Authority (FINMA), cryptocurrency companies would not require any specific approval or license to operate. |
|
Singapore |
The Monetary Authority of Singapore (MAS) stated that cryptocurrencies were assets, and the Authority would not regulate virtual currency transactions, but would monitor the relevant Know Your Client (KYC) and Anti Money Laundering (AML) processes. |
Source: David Yoffie and Anthony Woo, “Note on Blockchain and Bitcoin, 2017.” Synthesis by Yoffie and Woo based on Jonathan Keane, “The State of ICO Regulation? New Report Outlines Legal Status in 6 Nations,” CoinDesk, July 13, 2017, https://www.coindesk.com/state-ico-regulation-new-report-outlines-legal-status-6-nations/, accessed October 2017; and Stan Higgins, “Russia’s Central Bank Issues Warning on Cryptocurrencies and ICOs,” CoinDesk, September 5, 2017, https://www.coindesk.com/russias-central-bank-issues-warning-cryptocurrencies-icos/,
accessed October 2017.
22
Source: Goldman Sachs, “The (Lack of) Fear ‘Factor,’” August 8, 2017.
23
Exhibit 9 AirFox Blockchain Target Market (AirFox Analysis)
Source: AirFox, “AirToken (AIR): The token for mobile access,” white paper, August 21, 2017, https://www.airtoken.com/paper/AirFoxICO_WhitePaper_v1.5.pdf, accessed January 2018.
24
Exhibit 10 AirFox Target Customer Profiles (AirFox Write-Up)
Prepaid Mobile Subscriber in Indonesia Aditya uses IndoSat prepaid service and often has to pay over 10% of his income in mobile service, he uses a low-end android device. He does not own a computer and his smartphone is his primary method for payments, connectivity, and communication. For him, the smartphone is an essential tool for work and his daily life. Aditya is very aware of the cost of his mobile service, constantly switching between carriers, often multiple times per month by having multiple prepaid connections in order to maximize his value. With AirFox, Aditya can earn and purchase AIR in real-time via AirFox’s mobile apps. He is now earning AIR for simply using his phone, without much change to his smartphone behavior. He can then use these AIR to claim data recharges for free or a highly discounted rate.
Indian immigrant in the U.S sending remittances back to India Rohit is an Indian immigrant in the United States constantly sending money back home for his family. Remittance and exchange fees are often expensive. With AirFox, remittance becomes easier and cheaper. Rohit can send AIR that he has purchased, earned or micro-financed to anyone in the world with a prepaid phone or with an AirFox account simply by inputting their phone number, all without the high remittance fees. The fact that Rohit is in the U.S. means he can claim higher amounts of micro-loans in AIR than his peers in Indiana. However, he can send his loans of AIR across borders for no-fees and be responsible for those payments in the U.S. Imagine a $200 loan in the U.S. can equate to nearly 12,600 rupees in India.
Unbanked user in Mexico Juan Pablo is a hard working blue collar worker in a factory. He operates in an extra-legal economy—cash only. His employer pays him in cash and banks prefer not to serve high-risk customers like Juan Pablo; it is just too expensive. Juan Pablo does not mind since the fees are just too high to justify being a banked customer. However, being unbanked precludes him from many capital benefits, including access to leverage and interest rates. It also precludes Juan Pablo from ever making purchases online. With AirFox, Juan Pablo uses AIR as the institutionalized currency and his smartphone as the bank. He can claim micro-loans of AIR that can be used for emergency capital, investments, and purchases. And never before, across AirFox’s direct ecosystem, Juan Pablo can also purchase digital and physical products from his smartphone using AIR.
Underbanked worker in Brazil Carlos has a low credit score from bad loans that he has accrued over the years. He has this great idea for a small business—a new type of hot dog stand—yet he does not know how to find capital to fund it. He does not need much, he wants to sell his wife’s famous hot dogs on the beaches of Rio and thinks he could quickly repay it within 6 months once he is operating. Credit cards could be an option, but 358% yearly APRs are unfeasible. With AirFox, Carlos not only can earn capital via advertising but he can also qualify, based on his smartphone data, to new sources of capital via AIR. The more he uses AirFox, the more he can redeem. After he has become a loyal user, we will continue to offer micro-financing options, as long as he pays his debt in the appropriate time. Even better? AirFox rates can be up to 10x cheaper than local banks and not discriminatory against federal credit scores since AirFox uses its own data score to determine the loan rather than relying on archaic models.
Source: AirFox, “AirToken (AIR): The token for mobile access,” white paper, August 21, 2017, https://www.airtoken.com/paper/AirFoxICO_WhitePaper_v1.5.pdf, accessed January 2018.
25
Exhibit 11 AirToken Market Flow (AirFox Visualization)
Source: AirFox, “AirToken (AIR): The token for mobile access,” white paper, August 21, 2017, https://www.airtoken.com/paper/AirFoxICO_WhitePaper_v1.5.pdf, accessed January 2018.
26
Exhibit 12 Sample AirFox Android Apps Powered by the AirToken Blockchain (AirFox Write-Up)
Source: Company documents.
27
Endnotes
1 Phil Goldstein, “Will today’s next-generation MVNOs shake up the U.S. market?,” Fierce Wireless, April 3, 2011, https://www.fiercewireless.com/special-report/will-today-s-next-generation-mvnos-shake-up-u-s-market, accessed January 2018.
2 Anne Morris, “Report: Number of MVNOs exceeds 1,000 globally,” Fierce Wireless, September 2, 2015, https://www.fiercewireless.com/europe/report-number-mvnos-exceeds-1-000-globally, accessed January 2018.
3 Aaron Smith, “U.S. Smartphone Use in 2015, Chapter One: A Portrait of Smartphone Ownership,” Pew Research Center, April 1, 2015, http://www.pewinternet.org/2015/04/01/chapter-one-a-portrait-of-smartphone-ownership/, accessed January
2018.
4 Jon Russell, “Former Mozilla CEO raises $35M in under 30 seconds for his browser startup Brave,” TechCrunch, June 1, 2017, https://techcrunch.com/2017/06/01/brave-ico-35-million-30-seconds-brendan-eich/, accessed January 2018.
5 Heather Walker, “How Digital Signatures and Blockchains Can Work Together,” Cryptomathic (blog), October 3, 2016, https://www.cryptomathic.com/news-events/blog/how-digital-signatures-and-blockchains-can-work-together, accessed December 2017.
6 David Yoffie and Anthony Woo, “Note on Blockchain and Bitcoin, 2017,” HBS No. 718-433 (Boston: Harvard Business School Publishing, 2017), http://hbsp.harvard.edu, accessed December 2017.
7 Yoffie and Woo, “Note on Blockchain and Bitcoin, 2017.”
8 The Investor’s Field Guide, “Hash Power—A Documentary on Blockchains & Cryptocurrencies, Episode 1: Understanding Blockchains,” podcast, September 26, 2017, http://investorfieldguide.com/hashpower/, accessed December 2017.
9 John Plansky, Tim O’Donnell, and Kimberly Richards, “A Strategist’s Guide to Blockchain,” Strategy+ Business, January 11, 2016, https://www.strategy-business.com/article/A-Strategists-Guide-to-Blockchain?gko=0d586, accessed January 2018.
10 Steven Johnson, “Beyond the Bitcoin Bubble,” New York Times Magazine, January 16, 2018, https://www.nytimes.com/2018/01/16/magazine/beyond-the-bitcoin-bubble.html, accessed January 2018.
11 Don Tapscott and Alex Tapscott, Blockchain Revolution: How the Technology Behind Bitcoin Is Changing Money, Business, and the World (New York: Portfolio/Penguin, 2016).
12 Marco Iansiti and Karim R. Lakhani, “The Truth about Blockchain,” Harvard Business Review, January-February 2017, https://hbr.org/2017/01/the-truth-about-blockchain, accessed December 2017.
13 Antony Lewis, “A gentle introduction to blockchain technology,” Bits on Blocks (blog), September 9, 2015, https://bitsonblocks.net/2015/09/09/a-gentle-introduction-to-blockchain-technology/, accessed January 2018.
14 “Drawing the distinction between the uppercase ‘B’ and lowercase ‘b’ in Bitcoin,” Blockchain (blog), December 29, 2014, https://blog.blockchain.com/2014/12/29/drawing-the-distinction-between-the-uppercase-b-and-lowercase-b-in-bitcoin/, accessed January 2018.
15 Satoshi Nakamoto, “Bitcoin: A Peer-to-Peer Electronic Cash System,” Bitcoin, https://bitcoin.org/bitcoin.pdf, October 31, 2008, accessed January 2018.
16 Nakamoto, “Bitcoin.”
17 Nakamoto, “Bitcoin.”
18 Ramana Nanda, Robert White, and Alexey Tuzikov, “Blockchain, Cryptocurrencies and Digital Assets,” HBS No. 818-066 (Boston: Harvard Business School Publishing, 2017), http://hbsp.harvard.edu, accessed December 2017.
19 Nakamoto, “Bitcoin.”
20 Nakamoto, “Bitcoin.”
21 Nanda, White, and Tuzikov, “Blockchain, Cryptocurrencies and Digital Assets.”
22 Chris Mooney and Steven Mufson, “Why the bitcoin craze is using up so much energy,” Washington Post, December 19, 2017, https://www.washingtonpost.com/news/energy-environment/wp/2017/12/19/why-the-bitcoin-craze-is-using-up-so- much-energy/?utm_term=.fcd2b106160b, accessed January 2018.
28
23 Ryan Browne, “Big transaction fees are a problem for bitcoin—but there could be a solution,” CNBC, December 19, 2017, https://www.cnbc.com/2017/12/19/big-transactions-fees-are-a-problem-for-bitcoin.html, accessed January 2018.
24 Jan Vermeulen, “Bitcoin and Ethereum vs Visa and PayPal—Transactions per second,” My Broadband, April 22, 2017, https://mybroadband.co.za/news/banking/206742-bitcoin-and-ethereum-vs-visa-and-paypal-transactions-per-second.html, accessed January 2018.
25 Timothy B. Lee, “A brief history of Bitcoin hacks and frauds,” Ars Technica, December 5, 2017, https://arstechnica.com/tech-policy/2017/12/a-brief-history-of-bitcoin-hacks-and-frauds/, accessed January 2018.
26 Lee, “A brief history of Bitcoin hacks and frauds,” Ars Technica.
27 “How to Store Your Bitcoins,” CoinDesk, October 19, 2015, https://www.coindesk.com/information/how-to-store-your- bitcoins/, accessed January 2018.
28 Adam Ludwin, “How Anonymous is Bitcoin? A Backgrounder for Policymakers,” CoinDesk, January 25, 2015, https://www.coindesk.com/anonymous-bitcoin-backgrounder-policymakers/, accessed January 2018.
29 Grace Caffyn, “Bitcoin on the Dark Web: The Facts,” CoinDesk, September 23, 2015, https://www.coindesk.com/bitcoin- on-the-dark-web-the-facts/, accessed January 2018.
30 Sherisse Pham, “North Korea is trying to amass a bitcoin war chest,” CNNTech, September 12, 2015, http://money.cnn.com/2017/09/12/technology/north-korea-hackers-bitcoin/index.html?iid=EL, accessed January 2018.
31 The Investor’s Field Guide, “Hash Power—A Documentary on Blockchains & Cryptocurrencies, Episode 3: Funding, Forking, and a Creative Future,” October 10, 2017, podcast, http://investorfieldguide.com/hashpower/, accessed Dec. 2017.
32 Mitchell Weiss and Elena Corsi, “Bitfury: Blockchain for Government,” HBS No. 818-031 (Boston: Harvard Business School Publishing, 2017), http://hbsp.harvard.edu, accessed December 2017.
33 Mike Orcutt, “How Blockchain Is Kickstarting the Financial Lives of Refugees,” MIT Technology Review, September 5, 2017, https://www.technologyreview.com/s/608764/how-blockchain-is-kickstarting-the-financial-lives-of-refugees/, accessed January 2018.
34 Anna Irrera, “Accenture, Microsoft team up on blockchain-based digital ID network,” Reuters, June 19, 2017, https://www.reuters.com/article/us-microsoft-accenture-digitalid/accenture-microsoft-team-up-on-blockchain-based- digital-id-network-idUSKBN19A22B, accessed January 2018.
35 Jason Kelley, “Blockchain in demand: Breaking into the growing job market,” IBM Blockchain Blog, October 9, 2017, https://www.ibm.com/blogs/blockchain/2017/10/blockchain-in-demand-breaking-into-the-growing-job-market/, accessed January 2018.
36 IBM, “IBM Blockchain,” https://www.ibm.com/blockchain/, accessed January 2018.
37 Jemima Kelly, “Top banks and R3 build blockchain-based payments system,” Reuters, October 31, 2017, https://www.reuters.com/article/us-banks-blockchain-r3/top-banks-and-r3-build-blockchain-based-payments-system- idUSKBN1D00ZB, accessed January 2018.
38 LiteCoin, https://litecoin.org/, accessed January 2018.
39 Monero, https://getmonero.org/, accessed January 2018.
40 CoinDesk, “Who Created Ethereum?” https://www.coindesk.com/information/who-created-ethereum/, accessed January 2018.
41 Ethereum, https://ethereum.org/, accessed January 2018.
42 Antony Lewis, “A gentle introduction to smart contracts,” Bits on Blocks (blog), September 9, 2015, https://bitsonblocks.net/2016/02/01/a-gentle-introduction-to-smart-contracts/, accessed January 2018.
43 Primavera De Filippi, “What Blockchain Means for the Sharing Economy,” Harvard Business Review, March 15, 2017, https://hbr.org/2017/03/what-blockchain-means-for-the-sharing-economy, accessed January 2018.
44 Seline Jung, “Filecoin v. Sia, Storj & MaidSafe: The Crowded Push for Decentralized Storage,” Medium/Token Report, August 3, 2017, https://medium.com/tokenreport/filecoin-v-sia-storj-maidsafe-the-crowded-push-for-decentralized-storage- 7157eb5060c9, accessed January 2
29
45 Trent McConaghy, “Blockchain Infrastructure Landscape: A First Principles Framing,” Medium/BigChain DB Blog, July 15, 2017, https://blog.bigchaindb.com/blockchain-infrastructure-landscape-a-first-principles-framing-92cc5549bafe, accessed
January 2018.
46 The Investor’s Field Guide, “Hash Power, Episode 1.”
47 Ramana Nanda, Robert White, and Alexey Tuzikov, “Initial Coin Offerings,” HBS No. 818-067 (Boston: Harvard Business School Publishing, 2017), http://hbsp.harvard.edu, accessed December 2017.
48 Interview with a venture capitalist, December 14, 2017. 49 Nanda, White, and Tuzikov, “Initial Coin Offerings.” 50 Nanda, White, and Tuzikov, “Initial Coin Offerings.” 51 Nanda, White, and Tuzikov, “Initial Coin Offerings.”
52 Nick Tomaino, “The Token Economy,” Medium, March 14, 2017, https://thecontrol.co/the-token-economy-81becd26b9de, accessed January 2018.
53 Pete Rizzo, “Ethereum’s Creator Proves Blockchain Scaling Vision is No Joke,” CoinDesk, September 19, 2016, https://www.coindesk.com/ethereum-creator-vitalik-buterin-scaling-devcon2/, accessed January 2018.
54 Interview with a venture capitalist, December 14, 2017.
55 David Siegel, “Understanding The DAO Attack,” CoinDesk, June 25, 2016, https://www.coindesk.com/understanding- dao-hack-journalists/, accessed January 2018.
56 Siegel, “Understanding The DAO Attack.”
57 CoinMarketCap, “BitConnect,” https://coinmarketcap.com/currencies/bitconnect/, accessed January 2018.
58 “Ethereum’s Vitalik Buterin: Biggest Bitcoin investment platform likely is a Ponzi scheme,” The Next Web, November 3, 2017, https://thenextweb.com/insider/2017/11/03/ethereum-buterin-bitcoin-bitconnect/, accessed January 2018.
59 Stan Higgins, “BitConnect Investors Left in the Lurch as Token’s Price Drops 90%,” CoinDesk, January 17, 2018, https://www.coindesk.com/bitconnect-investors-left-lurch-tokens-price-drops-90/, accessed January 2018.
60 Interview with a venture capitalist, December 14, 2017.
61 Jeff Bussgang, “Summer of ICOs: VC Implications,” Medium, August 28, 2017, https://medium.com/startup-grind/the- summer-of-icos-vc-implications-ead720e8efdd, accessed January 2019.
62 Jay Clayton, “Statement on Cryptocurrencies and Initial Coin Offerings,” Securities and Exchange Commission, public statement, December 11, 2017, https://www.sec.gov/news/public-statement/statement-clayton-2017-12-11, accessed January 2018.
63 “SEC Announces Enforcement Initiatives to Combat Cyber-Based Threats and Protect Retail Investors,” Securities and Exchange Commission, press release, September 25, 2017, https://www.sec.gov/news/press-release/2017-176, accessed January 2018.
64 Don Weinland and Louise Lucas, “Asian investors flock to initial coin offerings despite bans,” Financial Times, November 14, 2017, https://www.ft.com/content/05b2748c-c903-11e7-aa33-c63fdc9b8c6c, accessed January 2018.
65 Arjun Kharpal, “Initial coin offerings have raised $1.2 billion and now surpass early stage VC funding,” CNBC, August 9, 2017, https://www.cnbc.com/2017/08/09/initial-coin-offerings-surpass-early-stage-venture-capital-funding.html, accessed
January 2018.
66AirFox, “AirToken (AIR): The token for mobile access,” white paper, August 21, 2017, https://www.airtoken.com/paper/AirFoxICO_WhitePaper_v1.5.pdf, accessed January 2018.
67AirFox, “AirToken (AIR): The token for mobile access.”
30
