To explore the question of where Amos should bank, we have created four scenarios to compare Amos’s bank fees at different stages of his digitization journey. In all scenarios, Amos’s monthly revenue is R45 000. He pays out R24 000 to his suppliers and R14 400 to his employees leaving him with R6 600 at the end of the month (see appendix A for all assumptions). In these scenarios, Amos withdraws cash from a participating retailer and deposits cash at a retailer or bank branch, depending on channel access.

In the first scenario, Amos’s business is entirely cash-based. While he doesn’t incur any bank fees, this scenario does present other potential costs, notably theft, which may occur less frequently and predictably than bank fees, but is likely to be significantly more catastrophic.

In the second scenario, the business is still mostly cash-based. Here, Amos’s only digital activity involves depositing his weekly profit into his bank account for safekeeping. Overall, this scenario is expensive for Amos as deposit fees are generally high. The cheapest option is TymeBank with total fees of R40, although he would have to take public transport to deposit cash at the nearest Pick n Pay or Boxer. The most expensive bank in this scenario is FNB which would cost Amos R264 per month. As noted, Bank Zero is not an option as the only cash amount you can deposit into your account is zero.

In the third scenario, Amos receives approximately half of his revenue in cash and the other half digitally. However, he still pays his suppliers and employees in cash, and he has to withdraw cash to make these payments. With TymeBank and FNB, Amos would not be charged for cash withdrawals at a retailer so he should choose a bank where the participating retailer is in walking distance from his business (i.e., FNB). Nedbank is the most expensive bank in this scenario with fees of R77 per month. Overall, fees are significantly lower than in Scenario 2 because he does not deposit cash.

In the final scenario, Amos’s business is almost entirely digitized. Most customers pay digitally, and he pays his suppliers and employees via EFT. Some customers still pay in cash – this amount is deposited into his bank account. Once again, TymeBank would be the cheapest option in this regard with Absa being the most expensive with fees of R660. This is driven by high EFT fees.

Graphic 1: Scenario summary

Curiously, Amos’s bank fees are lowest when half his revenue is in cash and where most of his payments are in cash (Scenario 3). In this scenario he does not have to deposit cash at all, and avoids the most expensive transaction, namely cash deposits. Where he receives 75% of his revenue digitally, and makes all his payments digitally (Scenario 4), he actually lands up paying higher fees, because he has to pay to deposit some cash into his bank account. Arguably, if the objective of the banking sector as a whole is to shift customers away from cash, deposit fees should be as low as possible, with fees loaded onto withdrawals. The difference in fees between the third and fourth scenarios (partial vs full digitised) and the high deposit fees associated with the second scenario might indicate that traditional banks do not appear to be particularly interested in digitizing small businesses, particularly township-based businesses.

The full cost of going digital

These scenarios only partially explore the fees of going digital, as they do not consider the cost to Amos of receiving digital payments. With a Yoco device, Amos would be charged 2.95% per transaction. Therefore, if he receives R22,500 in digital transactions (half of his monthly revenue), he will be charged R664.

While Amos does not track costs closely, he knows he is charged for accepting digital payments, and so he charges his customers an additional R5 for card payments. Despite this, Amos’s business now generates roughly 80% of its turnover digitally. His experience indicates a pent-up demand for digital in townships that might surprise many digital financial services experts.

The future of business banking in South Africa

Fee structures for small business banking solutions offered by larger banks do not appear to encourage clients to go cashless. At the same time, the larger banks have been happy to leave the small business payments market to Yoco. This is curious. Aside from reducing costs to serve, increased digitisation of small business clients improves the visibility of business activity thus facilitating more lending to these clients, which is surely the ultimate banking objective.

It might be time for big banks to rethink business banking entirely. Their traditional markets are not growing and their offerings are complex and expensive. At the same time, there is a real opportunity to create bankability within the small business segment of the market, particularly for those that operate in townships.

Amos’s case study illustrates that this is not an exercise in upliftment that requires donor support. With Amos currently paying close to R1000 per month in fees to banks and payment service providers, not to mention the potential to grant his business and his household credit, it is surprising that more banks don’t see this market opportunity.