Academic

Multi-Jurisdiction Financial Services – A New Global Perspective

/ DC / Leave a comment

This articles highlights the key research I conducted and contributed to a broader working paper during my tenure at MIT’s Media Lab in 2017. The full paper can be accessed here and examines the current alternative technological methodologies employed by credit providers (lenders) and intelligence providers (analysts), the limitations of their business models and challenges that they face, and then seeks to identify potential visions of the future for credit provision to consumers on a global scale. The full paper is reproduced with kind permission of co-authors Omosalewa Adeyemi and Raunak Mittal.

What Alternative Methods can be Used to Assess Creditworthiness, and What are the Barriers Preventing More Open Access to Lending?

Abstract

In 2014, the World Bank estimated that 2 billion adults in the world lacked access to a transaction account and are excluded from the formal financial system. In conjunction with public and private sector partners, the World Bank Group set a target to achieve ‘Universal Financial Access’ (UFA) by 2020. The goal of UFA is for adults globally to have access to a transaction account or electronic instrument to store money, send, and receive payments by 2020.

While the implementation of UFA would represent a significant step forward for low income and underbanked populations around the world, the enormous potential of mass-market consumers to drive economic growth in emerging countries has been barely tapped. Consumer financial services can help raise the low income population – in developing and developed economies – out of poverty, however there is a significant barrier to opening access to credit to the due to the high customer acquisition cost faced by traditional for-profit lenders.

In this paper, we look at the current alternative technological methodologies employed by credit providers (lenders) and intelligence providers (analysts), the limitations of their business models and challenges that they face, and then seek to identify potential visions of the future for credit provision to consumers on a global scale.

Introduction

Current credit provision solutions barely scratch the surface when it comes to addressing the needs of low-income and unbanked populations in developed and developing economies. While growth in developing economies has been happening faster than in developed economies, financial services at the individual consumer level are struggling to catch up. Despite the hype surrounding micro-finance in recent years, a large number of low-income communities still have no access to formal sources of credit.

The key barrier to fully opening access to credit to the poor and unbanked is the high customer acquisition cost faced by traditional for-profit lenders. Conducting background checks and adhering to “know your customer” (KYC) standards is labor intensive – due to a lack of customer information for risk assessment – and regulations in many countries require credit providers to undertake detailed customer identity verification even for small transactions[1].

Nonetheless, there exists an enormous potential market if banks and other financial institutions are able to embrace financial inclusion of the poor and underbanked. Despite the significant upfront costs and challenges, we argue that institutions should seek to harness this long-term potential – utilizing advances in technology and government stimuli – to offer not only payment and remittance solutions, but access to a wider range of financial products and services.

In this paper, we look at the current alternative technological methodologies employed by credit providers (lenders) and intelligence providers (analysts), the limitations of their business models and challenges that they face, and then seek to identify potential visions of the future for credit provision to consumers on a global scale.

Credit Providers

Roughly speaking, existing credit providers can be assessed along two axes: (x) for-profit vs. nonprofit and (y) local vs. multinational. Local for-profit companies operate in one country and have built efficient and (relatively) effective lending products within those markets, utilizing their experience to grow the business. One example, Branch (established in 2015), is based in the USA and Kenya and provides loans to individuals and small business owners based on algorithmic decision-making involving mobile phone data – such as GPS location, call/SMS history and patterns – and battery status. These loans range from $2.50 to $500[2] and require a mobile money account to receive funds and make repayments. In common with most credit providers, borrowers can build a credit profile based on their repayment history in order to access lower interest rates and/or larger loan sizes over time.

Branch faces competition from a number of similar companies (M-Shwari, Saida, Tala, to name a few). Tala (formerly InVenture), an example of a multinational for-profit company, offers credit through its app which claims to utilize over 10,000 data points on each customer’s (borrower’s) phone[3], from financial transactions to daily movements via GPS. Loans range in size from $10 to $500, with an average amount of $50, 11% interest rate and repayment rate of over 90%. To date, ~66% of its 30-day loans have been used for small business purposes[4]. Tala has a presence in the USA and Kenya, and operates throughout East Africa and Southeast Asia, in countries like Tanzania and the Philippines.

Despite the promise of expansion (scaling) across borders, the use of mobile phone data is still somewhat primitive and it remains to be seen if this data alone is a reliable enough indicator of creditworthiness to support a large commercial lending venture. There is also an argument that individuals that have access (i.e. credit means) to regularly use their mobile phones are more likely to favor borrowing from their family and friend network to avoid the high interest rates and strict repayment terms demanded by commercial credit providers.

Figure 1: Example of lending decision process (Source: Branch[5])

Nonprofit and multinational operator Kiva, on the other hand, is active in 83 countries (including the USA) and has provided credit to approximately 2.4M borrowers to date[6]. Unlike Branch and Tala – which have funded themselves through a venture capital backed model – Kiva raises its funding via crowdfunding, targeting philanthropists and social change enthusiasts. Kiva pioneered this model in 2005 and has facilitated approximately $970M of loans to date[7]. Needless to say, this model is not easily scalable on a commercial basis given the need to provide a competitive return to investors.

U.S.-based LendUp is a for-profit venture targeting Americans in the lowest income bracket. It estimates that over half the U.S. population (more than 150M people) has a FICO score below 680, an arbitrary barrier for credit approval within most banks. LendUp offers short-term loans (up to $400 for up to 30 days) at spreads of 15% per month[8] across 24 states[9], allowing borrowers to build a credit history and hence access lower interest rates. LendUp does not provide much detail about the “most technologically advanced credit platform” that they created, but is not the only machine learning algorithm-based lender active in the U.S. Stilt is a 2016 Y Combinator alumnus that is committed to providing access to credit to immigrants within the U.S., hence broadening access to non-U.S. citizens who are effectively locked out of the local credit market.

Unlike the payments space, which is arguably already highly commoditized and multinational in nature, credit provision is typically built on a local model with deep expertise of the market, hence we witness a significantly higher number of local for-profit players.

Figure 2: Position of credit providers researched during this project

One current project that could provide a positive roadmap for future credit provision in the developing world is a partnership between Branch and Uber in Kenya. Uber has an incentive to facilitate access to credit so that its drivers can borrow towards a car, and in return provides its drivers’ data to Branch to assess creditworthiness[10]. All a driver needs to do to access a loan (initially KSh 30,000 or ~USD300) is to complete a minimum of 500 trips and have a 4.6* rating on the Uber app. Starter loans are repayable within 6 months at a monthly rate equating to 1.2%. The combination of new data and a (relatively) low interest rate makes this a compelling case study for future collaboration between commercial and financial institutions.

Additional (and reliable) data sources such as Uber driver information represent an exciting development in how future credit scoring might occur. A key requirement for opening access to credit to a more competitive market will be enabling such data to be available to a wider audience, beyond the individual user case for which the dataset was originally created. Richer digital data – via sources such as mobile phone usage – that can be analysed and employed as an informal ‘credit indicator’ can reduce the complexity of creditworthiness assessment and improve banks’ abilities to deliver services to a wider market[11].

Research / Interviews Conducted – Credit Providers:

Intelligence Providers

While the availability of funds to lend is obviously a key requirement of an alternative credit provision model, the ability to make informed decisions about credit decisions – and, hence, provide a sustainable business model for commercial lenders – is perhaps the most critical part of the equation. By definition, ‘intelligence providers’ (assuming they are not also lenders) can scale their operations more easily across borders, and most typically work in several geographies, tailoring their product offering according to local requirements.

In this section we examine the current trends between these intelligence providers, broadly along four dimensions:

  1. Data sources – where does an intelligence provider find its information?
  2. Interface – how does a partner/consumer interact with the service?
  3. Partners – who are the end customers?
  4. Business model – how do the intelligence provider (and its partners) make money?

Data Sources

Mobile phone data remains a key source of alternative data for intelligence providers, especially in emerging markets. U.S.-based Cignifi provides credit and marketing scores for partners such as Telefonica in order to reach underserved population in developing countries. Additionally, social network data and location (GPS) data are more commonly being utilized. Stanford-spinoff Neener Analytics, for example, uses personality and behavior analysis looking at a consumer’s social media footprint to score financial risk for thin-file, no-file or challenging consumers (estimated to represent 35-40% of U.S. consumers)[12].

Harvard spin-off EFL Global started with a straightforward psychometric analysis design, but now includes behavioral games in its credit assessment product. Applicants are asked to conduct simulations such as allocating funds to their household budget, which enables EFL to develop deeper insights into financial behavior as well as helping to prevent fraudulent activity on its app.

David Shrier, Managing Director of MIT Connection Science[13], believes that psychometrics and social media analytics have so far proven to be an unreliable measure of creditworthiness for existing fintech startups. A CEO of new MIT spin-off, Distilled Analytics, Inc., Shrier is working with predictive models that are 30-50% better at credit analytics than existing bank methods. Evolving from the findings of Professor Alex Pentland’s[14] studies involving social physics, Distilled Analytics, Inc. is not restricted to analyzing one data source, but is looking to the future and how it can disentangle the many credit indicators which are to be discovered in the masses of data being restored to consumers.

Two recent developments give an insight into the opening up of data ownership and privacy in the U.S. and Europe. In March 2017, the U.S. Senate (subsequently approved by Congress) supported a resolution[15] that paves the way for Internet Service Providers (ISPs) to sell consumers’ browsing histories to third parties. Across the Atlantic, from May 1, 2018, subjects of the European Union will benefit from the introduction of the EU General Data Protection Regulation[16] (GDPR) which includes the right for consumers to obtain electronic copies of any data being held about them from all commercial enterprises within the expanded EU territories covered under the Act. This heralds a huge leap forward for Europeans to access and control the data that is available and being seen by third parties in their decision-making, including the assessment of creditworthiness.

Interface

Intelligence providers in general are using cutting edge tech (data analytics, machine learning, etc.) in their products, whereas smartphone proliferation and reliable Internet access are potential barriers for expanding the service in emerging markets. Neener Analytics is a fully web-based B2B (SaaS) offering, whereas EFL Global allows consumers to take the tests in a supervised environment (with local “innovation”, in India they have someone with a tablet and a scooter to fulfil this purpose) or online via a web app, for example, with scores then being delivered to financial institutions through their API. New York City-based First Access offers a more customizable credit scoring platform for lending institutions in emerging markets which is accessible through their web interface or API.

In summary, there is no common agreement about the most effective interface between consumers (borrowers) and intelligence providers – the preferred model is likely to be a reflection of the technological maturity of the markets in which borrowers are based.

Partners

Financial services companies are, unsurprisingly, the predominant customers of intelligence providers. Traditional banks, credit unions and fintech lenders are all invested in this space, as well as mobile network operators (MNOs), investment companies, traditional credit agencies such as Equifax and retail store chains looking to expand their credit offering to desirable applicants. In most cases, intelligence providers provide an additional layer of credit scoring for its clients, which can be customized over time to complement, and potentially replace, a credit provider’s existing risk scoring model(s).

Due to the different lending criteria and credit models across financial institutions, intelligence providers typically work with their proprietary model (not trusting any dependent variable data from other sources except for pure financial data) and then expand it to incorporate actual data from the host client. MNOs form an important link in the partnership chain, providing access to mobile phone data which is a key component of many credit intelligence algorithms. Partnerships therefore are truly a two-way street, with data provision and scoring capabilities being the main commodities.

Business Model

There is a definite split in how analytics are being monetized, with traditional access fees (per report request, like traditional credit agencies such as Equifax and Experian) being replaced with specific ‘consulting-style’ partnerships between an intelligence provider and e.g. a credit institution and MNO. This reflects the high degree of customization which occurs, as well as a desire to ensure close control over consumer data and risk scoring data (which is treated as a competitive advantage of a lending decision-maker).

This raises two key challenges which exist in the intelligence provider ecosystem:

  1. How is data ownership and privacy maintained while it is being shared between the various partners?

Based on responses from the intelligence providers we interviewed, there are two key findings. First of all, there will continue to be friction and challenges to overcome between the incumbent banks and financial institutions (with their outdated standards and infrastructure for data privacy) and the advanced (cloud based, distributed, etc.) tech world as long as fintech companies attempt to disrupt the marketplace in new and innovative ways. Second, for most intelligence providers that work across different geographies, there will be a lot of variability in the standards they need to satisfy within their customer base.

Typically, an intelligence provider owns the psychometric data that is created via the borrowers’ interactions with its platform, and the bank or financial institution owns their own data. The bank will send anonymized records that the intelligence provider matches with a non-PII (personally identifiable information) key that has been created on their side. Such a structure allows intelligence providers to work with banks and financial institutions in jurisdictions with more onerous data privacy laws (e.g. Mexico).

Some intelligence providers have been able to make exceptions for countries with very strict regulations, such as where no data can leave the country (e.g. Indonesia). We are aware of a number of such incidents, during which an intelligence provider will establish a totally separate instance of its technology stack in-country in order to comply with regulations. Needless to say, such a setup is likely to result in higher costs being passed to borrowers but does at least provide a workable solution which can be iterated and improved upon.

  1. How can a consumer’s score(s) be transferred across different lenders/credit providers to enable a truly cross-border solution?

Nova Credit claims its Nova Credit Passport[17] – constructed from credit information and credit proxies (such as cell phone billing receipts and records) – is a truly global solution for immigrants to passport their credit scores on all their moves. Partnering with credit unions and fintech lenders in nine countries[18] they aim to open up ~$600B market in new lending opportunities to this highly educated and high-earning customer segment.

EFL Global has a medium-term plan to allow borrowers to take their EFL scores to other institutions (in the same jurisdiction or across borders), however it is complicated as banks have different lending criteria and credit models which are uniquely catered to by EFL’s one-to-one consulting services, making a generic product less valuable to individual lenders.

Research / Interviews Conducted – Intelligence Providers

How Might the Future Look

One intelligence provider we interviewed is already working on chatbot technology to enable an “anthropomorphized credit agent” with better UX (to help build trust and get more accurate answers), dynamic calling (no need to download an app, which is important in many emerging markets with limited data capacity) that can integrate with existing platforms (e.g. via SMS). We also heard consistently that mobile operating networks (MONs) are “sitting on goldmines” given the data they have (calls, top up history, messaging frequency, etc.), hence are likely to become a powerhouse of credit scoring data in the future.

Governmental initiatives like GDRP and the proliferation of IoT devices in the home and wider community will contribute more and more data and place it in the hands of consumers. While the opening up of personal data will introduce profound consequences for how we are perceived in a wide variety of settings – a topic that digital reputation visionary, Michael Fertik, explores exhaustively in his book ‘The Reputation Economy’[19] – it also offers a unique ability for the financial services sector to reinvent itself.

New technologies are in the pipeline that promise access to the large numbers of low-income and unbanked global communities in the future digital financial services marketplace.

Future banks and financial institutions (or however else they may be named) will eschew a central bank data repository, easily compromised, in favor of a secure, encrypted distributed data system. Personal data stores not only permit better digital walleting, but also greater security around personal biometric data which is integral to a future bank’s security protocols[20].

The adoption of digital currencies and distributed ledger techniques serves to drive down the ingrained financial transaction costs inherent in the current banking system whilst mitigating operational risks, which will offer financial incentives to future lenders to include low-income and unbanked populations, thus promoting financial inclusion on a global scale.

We expect AI to play a central role in the mission to disentangle indicators of intent from the masses of data being restored to consumers. Shrier, again, believes AI will enable ‘data monetization agents’ that can analyze individual consumer data in real-time and sell insights to the highest bidder(s) – think about breaking a shoelace as you go for a jog and being shown four advertisements for replacements when you look at your communication device – in order to provide a customized and beneficial service to individual consumers.

Such developments could easily serve to widen the financial inclusion gap between developed and developing economies as long as returns for commercial lending ventures are higher in regions where access to credit is already abundant. This raises the question of where there is a stronger appetite to adopt revolutionary technologies like digital currencies and share personal data to a wider audience – arguably this is higher in markets where there is no workable alternative in place today.

In any case, formal governance mechanisms will become increasingly important in order to overcome trust issues and promote the adoption of emerging technology. Governments and regulators should also work to ensure that consumer financial services are growing in developing economies, such that financial institutions of the future can eradicate poverty and harness the long-term benefits of this enormous potential client market.

[1] World Economic Forum Insight Report, ‘Redefining the Emerging Market Opportunity’, 2012

[2] https://branch.co/how_we_work. Accessed May 15, 2017

[3] http://tala.co/about/. Accessed May 15, 2017

[4] https://medium.com/tala/the-future-of-finance-starts-with-trust-bfa79f05893a. Published February 22, 2017

[5] https://branch.co/how_we_work. Accessed May 15, 2017

[6] https://www.kiva.org/about. Accessed May 15, 2017

[7] https://www.kiva.org/about. Accessed May 15, 2017

[8] https://www.lendup.com/rates-and-notices. Accessed May 15, 2017

[9] https://www.lendup.com/faq. Accessed May 15, 2017

[10] http://www.techarena.co.ke/2016/11/18/uber-branch-partnership/

[11] World Economic Forum Insight Report, ‘Redefining the Emerging Market Opportunity’, 2012

[12] http://www.neeneranalytics.com/results.html. Accessed May 15, 2017

[13] http://connection.mit.edu/

[14] http://web.media.mit.edu/~sandy/

[15] Senate Joint Resolution 34 (H. Res. 230): https://www.congress.gov/bill/115th-congress/senate-joint-resolution/34

[16] http://gdrp.eu

[17] https://www.neednova.com/lenders.html. Accessed May 15, 2017

[18] https://www.neednova.com/about.html. Accessed May 15, 2017

[19] ‘The Reputation Economy: How to Optimize Your Digital Footprint in a World Where Your Reputation Is Your Most Valuable Asset’, Michael Fertik and David Thompson

[20] ‘Frontiers of Financial Technology: Expeditions in future commerce, from blockchain and digital banking to prediction markets and beyond’, Visionary Future publication, David Shrier and Alex Pentland, 2016

Blockchain as a Solution to Monetize Digital Content and Protect User Privacy

/ DC / Leave a comment

This post is an updated version of a paper I co-authored (and reproduced with kind permission of my co-authors, Jackie Atlas and Lisa Conn) for the Entrepreneurship without Borders course[1] at The Massachusetts Institute of Technology (MIT) circa December 2016. I am delighted that the companies/concepts we chose to analyze are still thriving today.

INTRODUCTION

Overview

We live in an increasingly connected world, a world in which we have access to more information at our fingertips than our ancestors had in a lifetime. This increase of information has resulted in expectations for personalized, compelling, and “experiential” content. The production and distribution of this personalized content in the digital era requires new, sustainable business models. It also necessitates consumption of user data – of which consumers, governments, and companies are increasingly protective.

Blockchain, which can eliminate the one central hub of data and store all personalized data on individual devices, introduces a new solution for balancing consumer concerns around privacy and corporate needs to monetize content.

This paper explores two main questions around content monetization and user privacy.

  1. If consumers want to have access to great content, and current business models cannot support the individuals and companies that produce it, what can be done to create a sustainable model for this market?
  2. When it comes to balancing personalized content and user privacy, can decentralized blockchain technology allow for consumers to have both?

From there, this paper establishes a framework for evaluating blockchain solutions in terms of sustainability, security, cost, and adoption. We apply this framework to analyze two promising technologies, Brave[2] and Enigma[3].

CURRENT DIGITAL MONETIZATION STRATEGY

Content Digitization

Over the past century, our attention has shifted from the newspaper to the radio to the television to the desktop, and we are currently fixated on a smartphone in the palm of our hand. With each change in medium, new business models have been developed to support the production and distribution of content. But as impressions (eyeballs) have migrated so rapidly from one screen to another, and the access points (webpages) are seemingly endless, companies have yet to capture all of the value that is being created in the digital era.

When content first started moving online, corporations’ reactions were to simply sit back and wait. Look for trends. Hope to grow users and eventually charge a premium similar to the network effect route to success that apps can have in the Silicon Valley of 2016. Even a property as highly trafficked and relevant as ESPN took upwards of three years to welcome advertisers and generate revenue for branded content in hopes of this ‘premiumization’ effect. When there was no billion dollar ‘a-ha’ moment, websites adopted the same model as their television predecessors. But with the exponential increase in content available on the internet, these same impressions (and thus, dollars) became spread astronomically thin.

Power of Data

With every website click or search performed, an individual’s digital footprint becomes more robust and unique. The data that companies like Google and Facebook collect from each of its users contains insights about that person’s interests, habits, friends, etc. This data is extremely powerful for advertisers and other agencies trying to understand consumer behaviors.

The media industry historically captured this value through broadly targeted demographics (i.e. people aged 18-49). Television data is only refined enough to know what percentage of a show’s viewership may be made up these demographic brackets, and networks will take this information and sell it in 30 second commercial increments to advertisers. Data is collected (generally) by the Nielsen[4] company, which incentivizes its 40,000 homes to allow them to do so transparently, i.e. everyone in a Nielsen home is constantly reminded that they are being tracked. Online, data is not being collected from just 40,000 homes (approximately 100,000 individuals), but across billions of people worldwide. And instead of knowing minimal demographic information on its viewers, websites have incredibly precise measures of who is interacting with their content as each computer is tracked with cookies. When a 40 year old male who has recently searched for golf clubs and consistently reads the Financial Times lands on a page, it is incredibly easy for other high-end advertisers to access this data and target them.

On the internet, there is an underlying awareness that one is being cookied as they browse, but there is clear evidence that points to people being generally uncomfortable with this concept. An estimated 150 to 200 million people use ad blockers on their desktop or laptop ad browsers and that number is growing at 41% a year[5]. The fact remains, however, that browsing data is incredibly powerful and is making companies like Google and Facebook billions of dollars. It is also allowing them to pointedly address individuals and more effectively distribute content, branded or otherwise. Addressable media is simultaneously exciting and alarming to consumers; individuals want content that is uniquely curated for their interests, but are increasingly conscious that their every click is being both monitored and monetized.

Advertising

Advertising has become splashed across websites in the form of banner ads, pop-ups, video pre-roll, and so on; there is a spot for it on a website and a cost attached to it (static viewing, scrolling over, click-through, etc.). But as quickly as companies attempted to capitalize on ad placement, consumers found ways to avoid them either out of fear of their privacy or simply because they are a nuisance. Migrating away from sites that run slower because of ad noise, installing ad blockers, and “click fraud” has sent clear signals to companies about how their website is perceived. And consumers’ desire for uninterrupted web browsing experiences has proved to be a major hurdle for corporations trying to monetize content on the internet. A handful of players with strong brands have seen minor success using subscription models and paywalls, but for the most part, the “free” choices available have proved television’s demographic-based model to be unsustainable.

In 2015, internet advertising revenues reached $59.6B, however the largest takers of this revenue are the tech companies and platforms that hold the consumer data, not the content producers. So, if advertising is the current mode of monetizing content and allowing the internet to be (more or less) financially viable, what tools and methodologies can be utilized to drive revenues to the sources of content? This is where the blockchain fits in. As targeted advertising enhances consumer awareness that their data is being tracked, the desire for privacy grows, and more people will move to ad blocking technologies, threatening all web platforms. The blockchain provides an opportunity to disaggregate data collection so that personal information is not the property of one specific company, instantly addressing privacy concerns. There are also blockchain enabled technologies which can put this data into the individual’s control, where they can choose how it is utilized and monetized. While companies like Google and Facebook would be resistant to this blockchain application, it becomes more and more feasible as people demand privacy.

BALANCING PERSONALIZATION AND PRIVACY

Overview

Even increasing numbers of consumers are demanding personalized, immersive, and customizable experiences—from the ads they see to the content they consume. We can define these “personalized experiences” as interactions with a piece of content or technology that leaves the consumer feeling like their interests and preferences were being taken into account. But this desire comes into sharp contrast with another trend: decrease in consumer trust and desire for privacy.

Personalized Content

Younger customers, through years of experience in the digital world, have grown accustomed to the way technology can reduce the need for human gatekeepers to ensure accuracy and manage data. Amazon, Netflix, Hulu, Spotify, and Pandora for instance have trained consumers to expect personalized recommendations based on their past purchases—or what they have already listened to and watched. Similarly, consumers are exposed to seemingly endless information and content, resulting in information overload. Personalized experiences help reduce the perception of information overload by increasing the sense of control[6].

Increase in Demand for Privacy

While opportunities for personalization increase, consumers are recognizing the danger and demanding privacy. A recent Microsoft survey found that 75% of people were concerned about online tracking and thought that the “do not track” feature should be turned on by default. A study conducted by Toluna[7] found that 72% of Americans did not want to purchase Google Glass because they were concerned that private data recorded could become public. And, while Snapchat has recently changed its privacy policy, resulting in backlash in the media, it experienced rapid growth through its claim that photographs were not stored and data was not recorded.

Can one technology serve all markets?

Not all consumers are the same. Youth and millennial consumers, many of whom grew up online, are willing to trade privacy for more personalized content and services. In a Pew Study about internet privacy in 2025[8], Niels Ole Finnemann, a professor and director of Netlab, DigHumLab in Denmark, said: “The citizens will divide between those who prefer convenience and those who prefer privacy.” The future of personalized content will require increased privacy protection, or the ability to take data, make it less sensitive, but still of value to the user.

BLOCKCHAIN ENABLED BUSINESS USE CASES

Enter the Blockchain

In the current environment, user data is increasingly centralized. The shift from desktop to mobile shows that people want to be able to access their data from any device. The current solution is a cloud-based server, which is vulnerable to hacks and abuse. Blockchain could eliminate the honeypot of data, storing small parcels of personalized data on individual devices. From there, consumers can choose what they want to share about their identity. We imagine a solution in which the data collected on the internet does not belong to anyone – no longer to Google and Facebook – but rather is decentralized on the blockchain.

In order to best evaluate blockchain-centered solutions for content monetization and user privacy, we must establish a framework for analysis.

  • Sustainable Business Model: Does this technology support the individuals and companies that need to produce content?
  • Privacy: Does this technology protect user data, as much as consumers want their data protected?
  • Cost: What financial and switching costs are associated for content producers and consumers? Do the benefits outweigh the costs?
  • Adoption: How feasible is adoption of this technology?

Let us discuss Brave and Enigma, two blockchain-centered solutions that balance consumers’ concerns around privacy, and corporations’ need to monetize content.

Brave

Brave Software, Inc. (‘Brave’) blazed onto the browser scene earlier this year with two offerings: to make browsing the Internet faster and safer by automatically blocking ads and trackers, and redefining the relationship between content viewers and publishers using micropayment (Bitcoin) technology. Their overall business model is simple – download their browser and choose to see ads that respect your privacy, or pay sites directly to view no ads – but their mode of implementation is by no means agreed or even fully developed yet.

Brave initially offered its users the ability to choose to see ads that respect privacy (using anonymous protocols, like Anonize, and not tracking pixels, to confirm impressions) with a negligible effect on loading performance. Following its successful $4.5MM seed raise, on September 1st this year Brave went a step further and introduced its Bitcoin based micropayment technology, Brave Payment, which it released (like all its code) on an open source[9] basis. The technology allows Brave users to experience no ads and instead pay sites tiny amounts of money directly by simply turning this functionality on in their preferences page.

Sustainable Business Model: When users elect to see ads, Brave splits the ad revenues 55% to the publishers, 15% to each of Brave, its users and ad partners. On the face of it, this does provide sufficient motivation for content providers to accept the new technology (assuming that any loss of ad revenue is more than covered by a greater audience reach), but that didn’t prevent Brave from being threatened with legal action[10] from some of the biggest news contents providers, citing its model of ad replacement as being “indistinguishable from a plan to steal our content to publish on your own website.” This is more related to the no ads model, for which proof of sustainability is harder to provide.

Privacy: Brave’s initial setup provides highly increased privacy for consumers whose data will be accessible by third parties, but in an anonymized format. In addition, Brave claims to block all forms of ‘malvertising’, redirecting browsing to https protocol, and blocking tracking pixels and tracking cookies.

Cost: There are several costs to the user: first, the behavior of downloading the Brave browser after being familiar with a different browser. Consumers have passwords, credit cards, and addresses saved on their existing browsers. Consumers have faith in the speed and accuracy of the search results currently offered on familiar browsers such as Google Chrome. And in order to use the features that make Brave compelling, users would have to adopt bitcoin as a payment option. These costs are not insignificant, but could be overcome if the product provided enough value. We are unconvinced that privacy concerns are great enough – and that current solutions like ad blockers are insufficient – to lead to wide adoption. In the digital age, advertisers have grown accustomed to being able to target individual personas and interests; Facebook, for instance, promises the ability to target almost down to the individual, providing benefits such as ability to beta test new products, get instantaneous user feedback on the kinds of consumers that want the product, and sell to directly to the people who want the product most. As a result, entire advertising agencies and marketing methodologies have been built with knowledge of advertising on Facebook, Google, etc. Switching to Brave would require retraining entire teams, which is inevitable with any new technology, but the benefit would have to outweigh the inconvenience.

Adoption: Brave chose to partner with existing providers BitGo (wallet functionality), Coinbase (Bitcoin purchase) and Private Internet Access (to make IP addresses) in order to deliver a viable product today. While these partnerships make the product technically feasible, they do not help Brave actually gain adoption with consumers and then, advertisers. Brave has a tough road ahead: Google Chrome has 54% of market share of browsers[11]. Chrome, Safari, and Firefox have relationships with devices that ensure those browsers are pre-installed. Brave would have to build the same relationships, and/or spend millions of dollars persuading consumers of privacy risks and the benefits of switching to a blockchain enabled browser.

Concluding Assessment of Brave

As a technical product, Brave has a lot to offer. However, its adoption and feasibility is dependent on consumers becoming increasingly concerned about privacy in the future, and advertisers having no choice but to move.

Enigma

One team at MIT’s Media Lab is busily working on a new technology called Enigma, a peer-to-peer network powered by the blockchain that allows different parties to store and run computations on data while keeping it completely private. According to the founders’ White Paper, “Enigma is a decentralized computation platform with guaranteed privacy. [Their] goal is to enable developers to build ‘privacy by design’, end-to-end decentralized applications, without a trusted third party”.

Our team interviewed Thomas Hardjono[12], Technical Director at the MIT Internet Trust Consortium. We discussed using Enigma to create a personal data store provider where a user’s total browsing history could be saved in nodes on a peer-to-peer network. This treasure trove of potential behavior could be provided to advertisers via an interactive API for them to run statistical queries and provide targeted ads, thus enabling users to monetize their data without it being stored or processed on the publisher’s servers.

Sustainable Business Model: This proposition is more attractive to content providers than Brave’s model, since the full customization of ads can be provided to web users with data about users’ interactions with ads being stored and available for follow-up queries. Enigma’s inventors say the technology is still several years from being available in a commercial format, so it’s difficult to evaluate the economics of the business model. This business model relies on consumers being willing to monetize their browser history – and would allow them to choose what content is shared.

Privacy: Using secure multi-party computations, queries can be run in a wholly distributed way, with data split between nodes on the blockchain and computations being run without sharing information with other nodes. Advertisers will never have access to data in its entirety, and are not able to run point queries (such as asking for users’ identities) but instead can run statistical queries about, e.g. users’ travel habits. Due to the introduction of a large number of nodes, the system is also highly resistant to losses stemming from hacking. For consumers, this solution provides flexibility—it allows individuals to share personal data to whatever degree they are comfortable.

Cost: If Enigma is utilized to allow consumers to monetize their browser history, it would represent a paradigm shift in the relationship between consumers, content providers and advertisers in the digital age. Since Enigma is not fleshed out in its implementation just yet, it is difficult to fully assess the behavioral or financial costs with any specificity.

Adoption: This technology is not yet proven. Even Hardjono described this use case as “using a sledgehammer to crack a nut.” Until the requirement for the blockchain to be a central solution for this issue is widely accepted, it may remain a pipe dream for the foreseeable future.

Concluding Assessment of Enigma

While the reality of Enigma is far off, the concept could provide consumers with the most control over their own data while allowing content producers to monetize their content with targeted advertising and information.

CONCLUSION

It is uncertain what the future holds regarding consumer’s interactions with online content as behaviors continue to evolve. The rate at which ad blockers are being adopted suggests that individuals are uncomfortable with the ads themselves, and that privacy has become a big enough concern to drive changes in a consumer’s interaction with the internet. As data is currently being aggregated by companies and other centralized owners, the blockchain is a feasible answer to solving the privacy concern portion of this question.

In this nascent stage of the blockchain, it can be argued that switching costs are too high and that digital currency is not widely enough accepted to drive drastic changes in any new platform’s acceptance. Additionally, we assume that people prefer personalized experiences over anything else and are willing to give up their data for better user experiences. Content producers would be eager to find ways to personalize (and monetize) their offerings, but this is not a strong argument to change behaviors of the end user. But in a world where people are genuinely concerned about their privacy, as we are undoubtedly moving towards, Brave’s blockchain solution is a winning one. When individuals have the opportunity to own their own data, protect it, and given the option to distribute it via the blockchain, Enigma’s platform is also highly compelling.

We can hypothesize that it will only take a few well publicized privacy hacks to drive enough awareness so that consumers are demanding changes in current offerings. The crux of decentralized privacy and personalized data offers digital platforms an opportunity to rethink how they are monetizing their content. It is impossible to tell whether that will happen through blockchain-enabled browsers, the individual’s choice to opt in/out or even be paid to view ads, or something far different that has yet to be developed, but the looming upside is undeniable.

[1] Also known as Implications of Blockchain Technology for Economic and Financial Development

[2] https://brave.com/

[3] https://web.media.mit.edu/~guyzys/data/enigma_full.pdf

[4] https://www.nielsen.com/us/en/about-us/panels/ratings-and-families/

[5] https://www.theguardian.com/technology/2016/jan/03/web-advertisers-blocking-digital-monitoring-ethan-zuckerman

[6] “Consumer Control and Customization in Online Environments,” Laura Francis Bright. University of Texas. https://repositories.lib.utexas.edu/handle/2152/18054  

[7] “7 out of 10 Americans will avoid Google Glass over privacy concerns,” Mike Flacy. Digital Trends. http://www.digitaltrends.com/mobile/7-10-americans-will-shun-google-glass-privacy-concerns/#ixzz4NHTQXAgz

[8] “Privacy in 2025: Experts’ Predictions,” Pew Research Center. http://www.pewinternet.org/2014/12/18/privacy-in-2025-experts-predictions/  

[9] https://github.com/brave

[10] “Publishers Seek to Stop Brave Browser Ad-Blocking Tool.” Wall Street Journal. http://www.wsj.com/articles/publishers-seek-to-stop-brave-browser-ad-blocking-tool-1460065209

[11] https://www.netmarketshare.com/browser-market- share.aspx

[12] https://hardjono.mit.edu/