Counting More than Accounts: What Can Frequency of Transactions Tell Us About Financial Inclusion?

Analyzing frequency of usage gives insight2impact an indication of how populations are (or aren’t) switching to digital channels for transacting.

In Nigeria and Mexico, cash is ubiquitous in daily transactions.

What counts as financial inclusion? Reporting the number of active accounts in a country is a common way to measure progress in financial inclusion. But as access to, and uptake of, accounts increases, we need more powerful indicators to help us understand the dynamics of how people use financial services, and digital financial services in particular.

At insight2impact, we’ve spent the last four years exploring these questions and have tested our measurement frameworks and emergent indicators on both demand-side survey data and transactional data in Mexico and Nigeria. We also went further by linking the data sets for the same person to analyze a more complete picture of people’s financial lives, both within and outside of formal financial services. What we found was both curious and transformative for our thinking on how we measure financial inclusion.

Frequency as a Means to Track Digitization

The first step to financial inclusion is having access to a transactional account. If I choose to take up an account, I’m able to store money or send and receive payments, but if I use my account very rarely, am I still financially included? Maybe I receive a payment and immediately withdraw it all at an ATM. What if all the transactions I need to do can only be done in cash?

Analyzing frequency of usage gives us an indication of how populations are (or aren’t) switching to digital channels for transacting. We analyzed bank transaction data in Mexico to better understand financial service usage patterns. Debit-card usage by customers was relatively high, with 78 percent using their cards weekly, but 60 percent of monthly card usage occurred at ATMs to withdraw cash, not for purchases via POS or online. The frequency of ATM vs non-ATM card usage paints a more realistic picture of digitization in Mexico. It’s a picture of incomplete digitization, with cash remaining central for most people.

Secondly, we linked the transaction data to survey responses to compare people’s online vs offline transaction behavior. This revealed that most people still use cash for their regular expenses, even if they have a bank account. The graph below shows both banked and unbanked people’s cash usage for regular expenses such as rent and bills, and daily expenses such as groceries and transport. Cash was ubiquitous, regardless of gender, socio-economic segment or account ownership. By only counting the accounts, we would miss the scale of non-digital financial lives.

Given the growing policy objectives of governments to go cashless, indicators that can measure the scale and nature of current non-digital payments and track the progress in digitization matter. Measuring frequency of usage and the main digital use cases may be the simplest way to do this.

Frequency as a Reflection of Functional Usage

We wanted to test if the frequency of sending or receiving payments digitally was a useful indicator of digital financial inclusion. If a person transacted at least weekly in this way, can we assume that this channel is used as a reliable, convenient and affordable way to transfer value?

We tested this question using transactional data from the Nigerian Inter-Bank Settlement System‘s (NIBSS) instant payment platform, NIP, and segmented consumers into weekly, monthly, less frequent and dormant users (see graph below). In this sample, 68 percent used the platform at least once a month, with 26 percent using it at least once a week. This low-cost system is linked to an account and can be accessed using USSD (a protocol that MNOs use to send automated reply messages back to users), mobile app or online.

From the demand-side survey data we learned that 45 percent of the respondents in Lagos and Kano were small businesses providing goods or services. They received more than half of their income daily, and they received it in cash (80 percent in Lagos and 96 percent in Kano). For population groupings that transact on a daily basis, measuring account activity once in 90 days (often used as a cutoff to define an account as active) misses the scale of their financial need to transfer value. To measure the outcome of financial service usage and digitization efforts, frequency can provide a more useful view.

Frequency for Digital Payments Use Cases

Frequency can also provide insight into the main use cases for digital financial services. From the demand-side survey data in Nigeria we observed that airtime and person-to-person transfers were by far the main uses. If people could also buy groceries or pay for transport using NIP or similar digital financial services, could this lead to a more cashless society? Unfortunately, the transactional data we had was not clearly tagged by type, so it was difficult to accurately identify the main use cases.

If we want to measure and understand the progress of digitization for payments, it is most useful to understand the use cases and the frequency of payments within these use cases. Where digital transactions are tagged correctly and consistently, this becomes much easier to do, but it is something not yet done, except for merchant codes for POS (though these could also be improved). The potential for better understanding their own customers may prove to be an incentive for financial service providers to improve the tagging of transactional data. Regulatory bodies may also have a role in requiring this.

Seeing Beyond Traditional Measures

As the financial inclusion sector matures and as some countries move closer to their Maya Declaration targets on financial inclusion, our thinking on how we measure success needs to evolve. Measuring account ownership and dormancy rates are still useful, but we need to complement this with more nuanced measures that can tell us about financial service usage and whether the service meets people’s financial needs. Using transactional data to measure the frequency of usage is one useful proxy indicator.

Transactional data to measure frequency of usage is a useful proxy indicator of financial service usage and whether the service meets people’s financial needs.

Understanding the use cases that are digitizing fastest is another. But until financial services actually meet needs, people will continue to transact offline. And that means looking beyond formal accounts to better understand customers’ broader financial lives.

For more information, read our Nigeria report and more about our Mexico pilot here (the forthcoming report will be published in partnership with La Comisión Nacional Bancaria y de Valores [CNBV]).

Join the Conversation

Stay informed. Subscribe to our newsletter.