Will Blockchain(s) Eat the Marketplace Stack?

Marketplaces are at the heart of Point Nine’s investment thesis, and have been since our inception. Some of our investments in this space include DaWanda, Delivery Hero, DocPlanner, and more recently StarOfService or Storefront. At the same time, we have been following closely the advent of blockchain technology, and led two investments in blockchain-based companies with Bitbond and Chainalysis. The aim of this post is to make use of our expertise in marketplaces to seize the potential of blockchain technology within the marketplace landscape.

In a nutshell, our theory so far is that it can impact marketplaces in 3 ways:

• First, a new generation of marketplaces based on blockchain technologies could emerge, and disrupt industries.
• Second, some new products based on blockchain technology could become an integral part of existing marketplace stacks and drive incremental business efficiency.
• Lastly, they could provide a solution to tamper marketplaces’ network effects and prevent marketplaces’ operators from extracting too much value from their users.

Platform thinking in the marketplace world

We thought a good way to start was the simple and very good framework created by Sangeet Paul Choudary in his book Platform Thinking. This framework provides a way to analyze any online platform, in this case a marketplace. Choudary views a marketplace as a combination of three main components:

(We’ll deconstruct this framework with some of the bricks that are interesting when we think about blockchain technology. The bricks mentioned below do not constitute an exhaustive description.)

A network of users

Buyers and sellers active in the marketplace. The role of any marketplace’s operator is to reduce friction in order to facilitate transactions within this network. There many actionable ways to reduce frictions, but 3 stand out as important pieces when we try to measure the impact of blockchain technology in the marketplace world:

1. Increasing the level of trust either by playing the role of a trusted third-party or by leveraging user reviews
2. Improving the matching between buyers and sellers
3. Improving marketplace’s UX and UI.

An infrastructure layer

The technical pieces that marketplace operators need to build to enable these transactions. These are not always visible to users, but it always includes:

1. A protocol to exchange value or a payment infrastructure (eg. Paypal or Stripe)
2. A protocol to exchange data (eg. TCP)
3. A hosting platform (most of the time AWS, Azure or another cloud service provider).

A data layer:

Any data stored on:

1. Users
2. Past transactions
3. Goods or services sold on the marketplaces.

The data layer fuels most current marketplaces’ efficiency. For example, a better and more granular understanding of each user allows marketplace operators to build a better matching experience. User reviews and historical transactions increase the level of trust. They are also the core of a marketplaces’ reputation systems.

Let’s quickly review Choudary’s three examples to make each of these pieces clearer:

• Craigslist stores almost no data on their users and relies on a very simple infrastructure. The whole value of Craigslist lies within its very large network of buyers and sellers.
• AirBnB in its early days was a case in point of a successful vertical marketplace. They leveraged a smaller community of people (hence a smaller network) interested in only one service: accommodation. The members of this community were matched more efficiently thanks to an improved technology infrastructure and a more granular understanding of their users based on data.
• Since then, they have done such a great job in reducing friction that they can now rely on a network that is probably as large as Craigslist’s and have stored an incredible amount of data on their users and their past transactions.

The social problem of marketplaces’ network effects

One of the reasons why investors like marketplace is network effects. From an investor perspective, investing in companies with network effects is great. Network effects create winner takes-all dynamic, and very often the biggest player in the market is way bigger than its competitors. So, as an investor, if you pick the winner, then there is a very high chance that you’ll meet high returns.

But, as Albert Wenger from USV puts it well in a recent podcast: “from a social and from an innovation perspective, there is a good and a bad angle to it. The good angle to it is, a network (here a marketplace) that connects everybody and provides incremental value to everybody is great. But, if there is a dominant network, then there is a legitimate question such as: What are the dynamics that keep that network from extracting too much value from the network of users and also from manipulating his network to its own advantage?”

The reality shows that we already see this happening in some of the dominant marketplaces. Some of Amazon’s sellers have for example seen their business going bankrupt because Amazon had decided that they were not sufficiently compliant to their Service Level Agreement. Others have been crowded out because Amazon was willing to sell directly higher-margin products. Overall, these sellers have gone bankrupt because they had built a business that depended entirely on a centralized marketplace, which is controlled by a commercial entity which business interest was not aligned with theirs.

The problem is that this should happen all the more frequently as online commerce grows, and, at the same place, as each of these marketplaces become ubiquitous.

If we go back to Albert Wenger: “We know very well that monopolies do not behave well because they are not checked by competition and so if we take the perspective of innovation which is what we invest in (as VCs), then markets that are dominated by very large players and have the ability to control what marketplace participant can do, at least to some degree, are not as great for innovation.”

To cut a long story short, an unobserved consequence of marketplaces’ network effects is that, the more these platforms grow, the more we rely on them, and the more we are dependent on them. The more these platforms grow, the less is our influence (or our bargaining power) as users versus the marketplace as a community (which is great), and versus the company operating it (which could be less great). If you want to know more, A16Z did a great a great job at explaining how network effects create moats, or barriers to exit.

Our role as VCs is also to identify the technologies that could challenge these economic equilibria that are not necessarily fair (and invest in the companies driving these changes). We’ve seen here how network effects could lead to unfair value extraction from marketplaces’ users. I’ll describe in the next paragraph how blockchain-powered solutions could tamper network effects and provide an answer to this evolution.

The intrinsic properties of blockchains and their impact on marketplaces’ stack

It’s now the right time to introduce blockchains, because one of the underlying promises of this technology is to bring the power back to the edges. In other words, give power back to users.

The framework detailed above will help us go through each of the marketplace’s components, and understand how blockchains are interesting from a marketplace operator’s perspective.

A P2P network of payment

The first and most obvious way by which blockchains could eat the marketplace stack is by changing the way we pay for goods in marketplaces. Paying with bitcoin, or any blockchain-enabled currency, can allow users to pay almost instantaneously, with very small transaction fees and without requiring any financial intermediary. This would completely change the overly long value chain of Internet payments (the front-end processor, …the acquiring bank…the card association…the payment processor…the issuing bank…the clearing house network (ACH) … this complexity being mainly driven by the fact that banks don’t work on the same database).

Self-enforceable (“smart”) contracts

Many vertical marketplaces have started growing once they’ve been able to act as a third-party guarantee, either holding the funds before the transaction is completed (escrow) and/or guaranteeing the quality of the goods/service sold through their marketplace. If you want to know more check Fabrice Grinda’s post: “The Evolution of Marketplaces”. This can be a cumbersome process for marketplaces operators, or at least drive significant costs.

Using cryptographic trust, which is at the core of blockchains’ smart contract capabilities, the conditions required for a transaction to happen, could be written in code. Funds’ release could also be unclenched automatically by predefined triggers written in code.

The simplest example of how automatic contract execution would work is a contract for sport live betting. Instead of having to rely on a third-party guaranteeing that the score is exact and then releasing funds, two players could sign a contract, mutually agreeing that the results provided by any Sports API would be considered as valid. Once results are available, this API would send the results as an input to a contract. This input would unclench automatically the release of funds. This seems a bit far fetched, and somehow specific to something where conditions are available on the internet. That said, the advent of the IoT could change this radically. Think of a container connected to the Internet, which arrival at a port could unclench automatically the release of funds between the importers and the exporters. Or check this video of what could be AirBnB’s next lock connected to the blockchain.

A protocol with built-in data

A protocol is a “specification of how entities should communicate”. TCP / IP is the decentralized protocol over which any internet application is built today. Bitcoin itself is a protocol to exchange value and data. Interestingly, the blockchain lets developers create new protocols on top of the bitcoin blockchain, or a based on a new blockchain, and customize it for a specific purpose. Colu’s coloured coin is a protocol built on top of the Bitcoin blockchain to issue and represent any assets digitally. LaZooZ is inventing a new protocol based on a new blockchain dedicated to real-time car sharing. Drivers are rewarded with appcoins (LaZooZ tokens) when they drive passengers in their cars. OpenBazaar is re-inventing peer-to-peer marketplaces with another protocol based on Ricardian contracts built on top of the bitcoin blockchain.

Individual ownership of data on identity and reputation

Blockchain technology allows users to store and decide at their will if they want to share their personal data with others. In other words, users’ data would not be stored on marketplace operator’s servers on AWS, but on an instance owned/managed by each user. More interestingly, we could own our personal data on one unique blockchain-enabled platform and use it on multiple marketplaces.

We would then be moving from a paradigm where each user would have a separate identity and reputation on each platform to a paradigm where users would have a single identity owned and controlled by himself (and stored in the blockchain).

This would:

1. Create a global network of trust,
2. Increase trust on each of the platform individually and
3. Prevent marketplaces’ ownership of our ability to transact.

Traity in Spain is already building such a platform in a centralized fashion and partners with several marketplaces. Blockstack is building the next generation decentralized authentication system. Consensys is building two very interesting projects with uPort and Repsys, which both work on the Ethereum network and let anybody manage their online identity and reputation. Authentq is finally a new company based in the UK, which sells a solution for Identity and Reputation management built on the bitcoin blockchain to existing centralized software products.

Trusted and authenticated reviews and reputation

User reviews have a dramatic impact on the level of trust on marketplaces. Academic research proves for example that one initial negative feedback from a buyer (justified or not) correlates with an average decrease of the growth of a specific seller’ weekly sales from +5% to — 8%. This partly explains why sellers could be tempted to manipulate reviews, post fake ones or delete others. The problem of reciprocal reviews is another of the challenges of marketplaces as they scale (never heard of the infamous “5 for 5” with a Uber driver at the end of the drive?). Using blockchain technology, any review could be authenticated, and solely users who had actually been through a prior transaction would be able to post reviews (for example by signing reviews with their own private key before posting them).

Decentralized hosting

Finally, it’s not only users’ individual data that could be hosted in a decentralized fashion but the whole application. When using OpenBazaar, users download a client and run the application on their private computer’s CPU. The whole network is maintained by users or nodes contributing their own computing power to the network. This means no down times, and a constant availability of the marketplace!

If we try to sum it up, we end up with the following evolution:

Full stack vs. technological enabler?

It’s clear that we are today at too early a stage to know which approach has the brightest future between being:

• A blockchain-based product part of existing centralized marketplaces’ stack and
• What we could pompously call a “full-stack” blockchain-enabled marketplace (another new term after the SaaS-enabled Marketplace :) ).

Each time we analyze a new blockchain-enabled marketplace, the question we wonder about is: can a new company derive sufficient competitive advantage from using blockchain technology to be more efficient than incumbents? Can Storj.io be better at selling storage in a decentralized fashion than Dropbox is with centralized servers? Can Golem be a better at selling computing power than AWS?

To illustrate this, let’s take two quick case studies of 1) blockchain as a technical-enabler to power existing marketplaces’ efficiency (here AirBnB) and 2) blockchain as a reason to build a completely new marketplace.

1. Blockchain and AirBnB?

In a recent report published in May 2016, Goldman Sachs’ research department shows that the use of blockchain could drive improvements for sharing economy marketplaces, and especially AirBnB, in three areas:

• Booking: building the necessary trust prior to a transaction is a complex process. AirBnB currently asks users to scan their ID and connect their social media pages. They also leverage user generated reviews to improve the level of trust. Using blockchains, the government ID of both parties would be authenticated once by a technical provider, made available to the counterpart at the will of each user, and travelers would be able to trust and rely on any reviews published on the host’s page.
• Payment: funds could be released per fulfillment of a smart contract written by AirBnB, and the payment credentials tied to an ID stored on the blockchain. Seeing even further, the state of a lock connected to the Internet could be one of the smart contract’s triggers in order for the funds to be released. Slock.it is currently developing a smart-lock, which could help in this case.
• Reviews: reviews would not be accepted unless they are digitally signed by a reviewer, who has been habilitated because he has in a transaction with the person about which he’s writing a review (“you can’t write a review if you haven’t paid for the service”). These reviews could also not disappear as blockchains are tamper proofs.

If you wonder about the impact of these improvements, Goldman Sachs estimates that it could drive an increase in booking fees for AirBnB by $3–9bn by 2020*, driving up AirBnB’s market share from 1.5% to 6.5% by this time. Interestingly, AirBnB bought Changecoin in March 2016 to evaluate these opportunities ;)

2. Is Storj.io the new Dropbox?

Storj.io is a new decentralized marketplace enabling anyone to lend or rent harddrive’s space online. We can view it as a marketplace for data storage. We all have free storage on our hard drive, or at least a hard drive that we don’t use. The whole idea of Storj.io is that by using this idle capacity, the company could build a more cost-efficient cloud storage solution. A solution cheaper for customers, which would also ensure total security, privacy and ownership of data at the same time. On Storj.io, data is sliced in tranches and stored securely between different nodes in the network. It’s also tamper-proof and only the owner of the data can re-unify the different slices stored in different locations using its private key.

By having a number of people contributing computing power to the network of connected hard drives, Storj.io can finally ensure that the service is never down. We’re not there yet, but if it met customers’ interest, Storj.io might have the potential to do to the like of Dropbox or Google Drive what AirBnB started to do to the hotel industry 10 years ago.

Challenges for VCs

One of the recent tweets published by one of the founders of Slock.it

Investing in blockchain related companies does not come without any challenges for VCs. In addition to the risks that we usually take (such as technology, competition, or market adoption risks), these new startups create new (interesting) challenges in our evaluation process. What is, for example, the defensibility or the long-term value of a marketplace which cannot rely on users’ lock-in, and therefore has lower/no network effects? What if because users own their own transaction history and reputation data they could switch to another platform seamlessly?

Brad Burnham shared USV’s investment thesis when investing in OB1, the commercial company linked to Open Bazaar. It is worth a read and clearly set this investment aside every traditional VC post “We’re investing in XX and here is why” ;)

Blockchain start-ups have also found innovative ways to fund their developments, be it by selling new coins on the blockchain they maintain or by creating a DAO. While very interested by these new initiatives, we tend to think that good VCs can provide additional value beyond their financial investment, at least we strive to do so :)

If you’re building the next generation of decentralized marketplaces or a game-changer technology for existing marketplaces, please get in touch with us, we’d love to have a chat!

A special thank you to Clement Vouillon, Pawel Chudzinski, Jonathan Userovici, Clément Jeanneau and Maud Coppey for reviewing the draft and feeding me with new ideas.