Danny Curran manages a lot of money. He is the head of fundraising and commercial services for the Irish Red Cross and is responsible for managing all of the donations that come in from the chapter’s 500,000 donors. In the digital age, online pledges should make that task considerably easier and more transparent. But not all of the Irish Red Cross’ donors use its online platform.

“2016 was the first year where online donations overtook regular donations,” Curran says. “People still send large amounts of money in the mail.” Receiving thousands of dollars by post is not unusual, he adds.

Those donors put a lot of faith in the Irish mail system. More than that, they put a lot of faith in the Irish Red Cross to spend their cash contributions on the pressing humanitarian causes that inspired them to donate to begin with, like refugee assistance or natural disaster relief.

States do that as well. In 2015, US$131 billion in development aid was committed by state and local governments worldwide, according to the Organization for Economic Cooperation and Development. But in any given year, about 30 percent of that money gets lost somewhere between the donors and the people and places it is intended to help, says Joseph Thompson, founder of U.K.-based software company AID:Tech.

Donor aid losses are a problem that Thompson first confronted on a charitable run across the Moroccan Sahara Desert. Donors had committed to his cause of helping children in need of reconstructive surgery, but it bothered him that he had no way of proving how those pledges were spent. “It got me thinking, what if someone made a donation and you could trace to whom and where that went,” he says.

That question prompted the launch of AID:Tech, a software company that tracks financial transactions, like aid donations, grants, or investments, from funder to the end point of use. AID:Tech does this by logging the flow of every commitment on a blockchain, a decentralized online ledger that records transactions.

Financial transparency is at the core of AID:Tech’s model, but its particular platform allows aid organizations to trace the disbursement of anything from remittance payments to credit vouchers to relief goods to doctors’ visits. “You can send anything through a blockchain,” Thompson says. (A digital representation of anything, that is.) And once a payment or item is sent, the transaction can be traced, but not edited or altered.

In late 2015, AID:Tech put that concept to the test when it teamed up with Curran and the Irish Red Cross to distribute relief funds for Syrian refugees living in Tripoli, in northern Lebanon. Ten thousand dollars was loaded across 500 “smart vouchers” coded with QR codes and linked to AID:Tech’s blockchain-based platform. Syrian families took the vouchers to a participating grocery store in Tripoli to buy food and home supplies. The moment the vouchers were scanned, the transactions showed up on Curran’s computer in Dublin. “I was watching the transactions come through in real time. It was amazing,” Curran says.

Following that successful test of AID:Tech’s system, the Irish Red Cross is moving forward with a bigger roll-out of the software across its donor platform. AID:Tech, meanwhile, has a number of new use cases in development with other non-profit partners, including a remittance payment system with the U.N. Development Programme (UNDP) in Serbia.

What exactly is blockchain?

Blockchain has been getting a lot of media attention as of late, thanks to a recent surge in “blockchain-based solutions” promising everything from specialized digital currencies, digital identity verification, supply-chain tracking—even election validation. The technology is still most often associated with bitcoin, the digital currency that blockchain was designed to support. Now, however, advocates posit that blockchain is foremost a transparency tool, aligned with an increasingly digital global economy.

When bitcoin was launched in the aftermath of the global banking crisis in 2008, the anonymous founder of bitcoin designed the underlying trading platform as a means of conducting digital financial transactions without the need for intermediaries, like brokers or banks. (Presumably, the idea was to create an alternative financial system—one that would give individuals more control over their financial assets while shielding them from potentially risky bank behavior.) The underlying platform, blockchain, depends on software just like other back-end banking systems; however, it is maintained and managed directly by the parties who use it.

“It is useful to think of a blockchain as a database that incrementally gets built up by a network of participating parties who run the same software.”

“It is useful to think of a blockchain as a database that incrementally gets built up by a network of participating parties who run the same software,” writes business and finance journalist Brett Scott in a recent working paper for the U.N. Research Institute for Social Development.

“A blockchain, as the name suggests, gets built up by blocks of data gradually being ‘chained’ together.” Each “block” of data captures one step in a transaction; each step is captured sequentially in a chain of blocks. “A blockchain database continues to be built and maintained so long as the software continues to be run,” Scott continues. “Thus, unlike a centralized database held by a single entity, it continues to stay ‘alive’ even if individual participants pull out.”

Because blockchain databases are built up gradually as transactions unfold, the system depends on dynamic memory, meaning it has to be able to add more storage space as new data comes in. Every time a new block of data arrives, the system is programmed to log the data point and then designate memory and a storage location for the next block. To ensure all of the blocks in the database stay connected, each individual block is programmed with its own storage location and the locations of the previous and subsequent blocks. (For the non-software engineers, this database structure is a type of “linked list”.)

Linked lists are not unique to blockchain. What makes blockchain’s application of a linked list unique is that the trail of data runs bi-directionally, meaning that from any given data point, the database’s users can follow the chain of data to all of the data points logged before and after. In a simpler linked list, users can only follow the sequence of data in one direction.

What’s more, all blockchain data is encoded with a random string of numbers and letters, also known as a “hash function.” Hash functions are what ensure a blockchain’s security. Each block’s hash function is linked to the subsequent block’s hash function; therefore, altering the data in one block would affect every other block in the chain. “It creates an indelible record, resistant to tampering by any individual party,” Scott writes.

Public versus private

The original bitcoin blockchain was developed as an open, public platform, unlike traditional financial trading platforms. Today, there are multiple blockchain platforms, including private, permission-based versions. AID:Tech uses the latter, so it can customize its software to meet different aid organizations’ needs.

“A public chain is a single global chain that everyone shares, and it’s visible to all of its users,” explains Flavien Charlon, AID:Tech’s technology lead. “Public chains are good for censorship resistance, or in places with [tight financial] controls. But for our clients, a permission-based version is a better choice.” That is because making aid organizations’ transaction-by-transaction data completely open to the public is unnecessary to ensuring that their finances are still traceable.

What is more important, particularly in times of crisis, is that organizations can efficiently manage and track the many different stakeholders and resources involved in their programs. “It’s impossible to know what would happen with a more opaque system, like cash distribution. With blockchain, every transaction is accounted for, and it is easy to drill down into the details,” Charlon says.

Curran witnessed that early into the Irish Red Cross’ pilot project in Lebanon. Within hours of distributing the smart vouchers, the participating shop owner started noticing error readings when he scanned some of the cards. “People started coming into the shop with fraudulent vouchers that they produced on the street,” recounts Niall Dennehy, who heads AID:Tech’s operations. “They were perfect copies, but the code couldn’t be duplicated. That’s how the cards were rejected.”

During the pilot, Dennehy says the shop owner revealed that he had participated in other refugee aid programs that distributed paper vouchers. He unknowingly cashed out thousands of dollars from fake vouchers in one case.

Curran says the pilot helped convince him of blockchain’s potential for improving financial transparency in the charity world. His hope is that using AID:Tech’s platform will boost fundraising for his organization. “The fundraising side is what excites me,” he says. “When members of the public make donations, [most of the time] they don’t know who and how it benefits. The more they know, the more they give.”

The pilot also shed light on the kind of support the Irish Red Cross’ aid recipients need. Being able to track purchases made with the smart vouchers will help the Irish Red Cross manage its logistics and relief preparedness better, both overseas and at home.

“We have a commitment to help about 4,000 refugees and asylum seekers resettle [in Ireland] per year, but when they get here, we don’t always know what they need [in regards to] clothes, hygiene, and food,” Curran says.

For example, Curran noticed that the families in Lebanon used the smart vouchers to buy household cleaning supplies, along with food and other necessities. Purchasing cleaning products would not be permitted under other relief organizations’ voucher schemes, because their aid programs are often restricted to food items only. (They often impose these restrictions to ensure vouchers are not used to buy things like alcohol or tobacco.) The Irish Red Cross-AID:Tech study’s feedback session exposed how limiting such restrictions can be to aid recipients, however.

“With my other vouchers I could not buy what I needed,” one participant said. “Only certain products were allowed, and others weren’t. We were not even allowed buy cleaning products or detergents, when we lived in places that need the most cleaning and disinfecting.”

Digital assets, digital identities

The benefit of permission-based platforms is that they are highly customizable; AID:Tech can tailor its software to any client’s need. Whereas the Irish Red Cross is primarily concerned with the traceability of its donor pledges, other organizations have different priorities. Healthcare providers offering clinical exams or immunizations, for instance, may want to record delivery of services; disaster relief responders may be concerned with how many temporary shelters or bottles of water they are distributing in a disaster zone.

Any “asset”, tangible or intangible, can be tracked digitally through a blockchain by encoding and logging it on the platform, Dennehy says. That same encoding can also be used for registering recipients of aid or social benefits. “You can take a person’s picture, capture their details, and embed that in the blockchain,” Dennehy adds. Entitlements can then be assigned to that person’s digital identity, whether it is 100 bottles of water from a relief organization or social welfare benefits from a government agency.

Unlocking the benefits of blockchain depends on access to personal data. Proponents and critics are interrogating how traceable data is being used and protected.

AID:Tech is currently working with UNDP on a program related to social welfare and remittances in Serbia. Remittances are a $600 billion global market, with the overwhelming majority of payments flowing from wealthy countries to low- and middle-income countries. (In Serbia, it is estimated that about 800,000 people receive $2.4 billion in remittance payments from abroad each year, according to estimates from UNDP.) Services like MoneyGram and Western Union dominate the money transfer market, but they are expensive, costing senders nearly eight percent per transaction on average. UNDP wants to bring the transactional costs of remittance payments down, while simultaneously helping countries like Serbia understand where these lucrative flows of money are going.

“People don’t always spend the money the way it is meant to be spent. Utility bills and education fees go unpaid,” Thompson says.

The solution AID:Tech is working on for UNDP will allow the organization to act as an intermediary between remittance senders and receivers. By partnering directly with the Serbian government’s welfare agency and local utilities, UNDP will offer a type of smart voucher for specific goods and services, like utility payments. A family member sending money home could then opt to send regular, undesignated funds—cash, basically—or funds encoded for a specific use via UNDP’s service. In either case, how the payments are spent would be fully traceable.

AID:Tech’s first remittance pilot with UNDP is getting underway with 200 Serbian social welfare beneficiaries. If it goes well, the platform could be rolled out across 21 other UNDP sites.

Proponents and critics of blockchain alike are interrogating how traceable data is being used and protected. “It would be a mistake to rush headlong into blockchain innovation without understanding how it is likely to take hold,” Harvard Business School professors Marco Iansiti and Karim Lakhani warn in a recent article in the Harvard Business Review. “Our experience studying technological innovation tells us that if there’s to be a blockchain revolution, many barriers—technological, governance, organizational, and even societal—will have to fall.”

Thompson explains that he is aware of critics’ concerns, particularly around data privacy. In AID:Tech’s case, he says, the company is just providing a software platform; it does not have access to its clients’ data. Instead, an organization like the Irish Red Cross or UNDP has full control over its own data, and can choose what information it does and does not share with its partners.

Dennehy says that he believes the social impact potential of digital identities outweighs privacy concerns. “There are two billion people worldwide without access to financial services. If we can give them a digital identity, they have a financial fingerprint that goes everywhere they go. They can use it to purchase micro-insurance [or secure] microloans,” he says, adding that this also empowers highly transient individuals like refugees, who often have or carry no documentation with them.

Tracking the impact

Indeed, promoting broad-based financial inclusion is one of governments’ and non-governmental organizations’ top priorities, particularly as they strive to meet the U.N.’s 2030 Sustainable Development Goals for poverty alleviation and economic opportunity.

There are challenges to scaling blockchain technology, however. For one, the original bitcoin blockchain is energy intensive, and requires both powerful servers and considerable data storage space to run. In many of AID:Tech’s target markets, such a platform would not work, because energy and communications technology infrastructure is often limited and/or unreliable. Thompson explains that AID:Tech’s permission-based platform is far less energy intensive than bitcoin’s platform, because the system is managed and maintained by one or two parties, not thousands of “miners” spread around the world. Thus, for most of AID:Tech’s clients, simple, low-data, low-bandwidth Raspberry Pi devices are sufficient for logging and tracking transactions. “The Raspberry Pis cost $8 to $10 per machine per year and they back up all of the transactions on the platform,” Thompson says.

“The most impactful distributed ledger technology applications will require deep collaboration between incumbents, innovators and regulators, adding complexity and delaying implementation.”

AID:Tech has also been in negotiations with a satellite provider to enable connectivity in extremely remote or low-infrastructure locales, like refugee camps. “With a satellite provider, an aid organization or government can ensure that every transaction happens in real time,” Thompson adds.

Physical infrastructure limitations are one issue; AID:Tech also has to contend with the slow-moving bureaucracies of large non-governmental organizations and government agencies. Because its projects often involve many stakeholders, educating all of the relevant parties about its technology and getting them signed on can be a lengthy process. Thompson says its conversion cycle for any given project can be anywhere from six to 18 months. Over time, it expects this will decrease, as clients replicate projects in new markets. Nevertheless, it is a reality the company has to contend with for now, as it lays both the commercial and social impact foundations of its business.

“It’s a lot of work to get all of the moving parts [working] together,” says Constantin Gurdgiev, a finance professor at the Middlebury Institute of International Studies in Monterey, Calif. Gurdgiev served as an early partner to AID:Tech as the company was starting to think through the social impact aspect of its platform. One of the projects he undertook on AID:Tech’s behalf was recruiting Middlebury students to identify key impact metrics for the company to track, and then plot out a potential scoring system for each metric.

Gurdgiev highlights an example of the students’ work from AID:Tech’s remittance project, where “transaction duration” was identified as a key metric for the company to track. “Often remittances are measured from the time a [payment] is sent to the time it is received, when it should actually be measured to when the money is used,” he explains. “It takes time to pay out of a senders’ account, then there are fees that have to be paid, then [the receiver] has to take the money to a payment center.”

When all of these steps are factored into the transaction time, the average payment is pretty inefficient, Gurdgiev says, “whereas with AID:Tech, someone could theoretically make a payment from the point of service.”

Gurdgiev is a big believer in blockchain’s capacity to drive efficiency and transparency in the financial markets, and calls it a game-changer for measuring social impact. “One of NGOs’ biggest problems is securing timely information for their donors to [prove] money isn’t being wasted,” he says. The real-time nature of blockchain-based transactions strengthens’ organizations quantitative data tracking, and gives them more robust data for identifying and contextualizing trends.

Gurdgiev acknowledges, however, that given AID:Tech’s target clients and the types of programs they run—with numerous stakeholders, most of whom are still unfamiliar with how blockchain works—the ramp-up time could be slow going. “It’s especially challenging on the bureaucracy side, which is why AID:Tech has to partner with agencies like UNDP, which can talk directly to governments,” he says.

This is hardly a reflection of AID:Tech’s individual model, however. “The most impactful distributed ledger technology applications [like blockchain] will require deep collaboration between incumbents, innovators and regulators, adding complexity and delaying implementation,” notes a 2016 report on distributed ledger technology by the World Economic Forum.

AID:Tech’s team says that its goal is not to overhaul or replace existing transaction-based systems; it wants to improve the traceability and accountability of them. “[Blockchain] is a way to create transparency,” Charlon says. “But it’s not a silver bullet that will solve every challenge.”