Iterating to Innovate: What We Learned with mVisa in Rwanda

The future doesn’t come with an owner’s manual saying how to set up, operate, or troubleshoot it. When we launched mVisa in Rwanda in 2013, it was the first interoperable mobile phone-based payment ecosystem in any emerging market. We didn’t know what was possible. But we knew what we were aiming at. We wanted to make mobile money work better.

Nearly all mobile money schemes are “closed loops”. They do not permit funds to be shared with users of any other scheme. Since consumers cannot transact with everyone they want or spend everywhere they go, they see mobile accounts as less useful than cash. Fewer make the switch from cash, the net financial inclusion impact is stunted, and commercial returns are blunted. The idea of mVisa is to connect the closed loops by routing mobile money transactions via VisaNet, the global software and data centers that process transactions by more than 2 billion account holders and sustain more than 30 million points of access in the Visa network.

We chose Rwanda to pilot the mVisa concept. A smaller market makes it easier to know and be known by key stakeholders. That is an important consideration when starting a multiparty ecosystem that requires all players to move in a similar direction in a similar timeframe. Rwanda fit the bill well.

How did we do? We didn’t get as many Rwandan customers as in our best case scenario. Three banks opened 150,000 mVisa accounts. But active use rates are on par with those of the 14 mobile money “sprinters” picked by the GSMA as industry leaders. Our Rwanda business doubled in less than a year. And the experience of having a team on the ground for two years yielded valuable confirmation of what worked while challenging us to evolve our approach on other fronts. We recently announced mVisa’s launch in India, with an updated strategy.

We’ve distilled some of our experience in Rwanda into 3 lessons:

1. Both ends of a payment ecosystem must grow at the same time. A payment system needs a payee and a payor. Otherwise, there are accounts but nowhere to use them, or vice versa. When they have been successful, MNO-led mobile money services have solved this in two ways. First, they began with remittances. Remittance senders and receivers have strong incentives to join the same service at the same time. Second, they seeded the ecosystem with payors via their own networks. When telcos sell airtime using mobile money, they themselves step in as one end of the ecosystem — a merchant. But few telcos have had success with any other use cases.

When we launched mVisa, these two use cases – mobile P2P and airtime sales – were already scaled by Rwandan telcos. We sought other opportunities. The largest microfinance bank, UOB, put more than half its microloan accounts on mVisa. A regional bank, I&M, partnered with the World Food Programme to pay monthly stipends in two of the country’s largest refugee camps via mVisa. Bank of Kigali, the largest Rwandan bank by assets, distributed a portion of the government’s US$19M in agricultural subsidies via mVisa accounts.

Each initiative featured a critical mass of consumers in specific geographies, enabling mVisa financial institutions to grow matching agent and merchant networks in the same locale. In doing so, we avoided most of the “chicken and egg” problem which bedevils many mobile ecosystems, particularly those that try to go nationwide all at once.

2. Consumer trust is earned, and the first stop is technical reliability. We were confident about the reliability of some of the “pipes” that mVisa relied upon. VisaNet hasn’t experienced a second of downtime in 21 years during the peak Christmas shopping season. We also felt good about launching mobile “push” payments for the first time in emerging markets. Globally, the vast bulk of payments involve a consumer sharing their account details with a merchant, typically by swiping their card on a card-reading terminal. Those details are then shared through the value chain to the consumer’s bank, which authorizes the requisite funds to be “pulled” from the consumer’s account to the merchant’s. The process matched available technology when originally devised, but with tradeoffs. A “push” payment reverses the flow, with the consumer always the one to authorize the transaction. This gives the consumer more transparency and control, cuts out a number of steps where consumer account details could be stolen, and reduces overall cost to process, which can help bolster the financial inclusion business case.

The technical challenges we experienced lay in the space between the consumer handset and VisaNet. VMMS, a new Visa service enabling the mobile channel for financial service providers, encountered long development timelines, particularly with integration to banks’ core banking systems. In new mVisa markets we are not selling VMMS with mVisa and we are putting greater emphasis on technical readiness of participating financial institutions.

mVisa clients also experienced a fair number of outages (and high cost) with the USSD and SMS channels managed by MNO networks. However, this isn’t a Rwanda-specific problem: the government of Zimbabwe has investigated a major MNO for competition practices.

3. The case for interoperability is still strong, but contextual.

How interoperability looks depends on who you are. When asked, consumers almost uniformly favor the ability to transact with more people and more places. Merchants want to be able to easily accept all forms of payment. Policymakers know interoperability increases market efficiency. And most financial service providers prefer it, too, but not all, including MNOs in Rwanda that declined to join mVisa. Particularly at early stages of market development, institutions that have captured a first mover advantage in a new technology often view interoperability as shrinking the gap with their nearest competitive followers. We’ve seen this with ATMs in the 1970s and 80s, and POS terminals more recently. Now, in the mobile money space, MNOs are generally turning aside offers of broad interoperability in favor of bilateral tie ups, which may offer them more negotiating power, but at the expense of creating a “spaghetti bowl” of different arrangements that are hard to navigate for consumers, and sometimes uneconomical for financial institutions to advertise except to a few well-heeled consumers. In recent CGAP research in Rwanda, 2 percent of Rwandans said they have sent or received money between different types of mobile money wallets and bank accounts.

Eventually we expect interoperability to be the norm in mobile payments, as it already is in other digital payment forms. mVisa is an attempt to bring that to fruition faster. Other steps are needed. The mobile channel needs to be more reliable, economical, and accessible to non-MNOs so innovation arises from all points on the compass. The shift from analog to digital unlocks huge potential for novel consumer-facing applications that drive uptake. Governments should up their usage of digital payments, not only for their own cost savings but also to add their weight behind the success of digital ecosystems.

Our next step is to take lessons learned from Rwanda and apply them in new markets, including in India, where mVisa launched in September. No doubt, we have more to learn there, too. In that sense, Visa is iterating in order to innovate.

Have you read? 

The Hold-Up with Mobile Money in Nigeria

“It All Comes Down to Knowledge” – Akiba’s Mobile Banking Experience

Ubiquity, Safety, and Utility: The Trifecta of Success in Financial Inclusion

Join the Conversation

Stay informed. Subscribe to our newsletter.