The apps that charge no fee to send money home are not free

Yes, the zero-fee claim is real. Of the fifteen services in our diaspora-friendly banks dataset, three advertise no transfer fee at all on African corridors. LemFi sends free to 10+ African countries and says openly that it earns from a small margin on the exchange rate instead. Taptap Send charges zero fees on key corridors including Nigeria 🇳🇬, with 95% of transfers arriving near-instantly. Sendwave runs low or zero upfront fees into West and East African mobile wallets.

None of that makes the transfer free. It moves the price from a line you can see to one you have to work out. The number that matters is not the fee, it is how much money reaches the person at the other end. Send the same amount through two apps on the same afternoon, write down the two figures the recipient would get, and you have your answer. Everything else is marketing.

Where the money actually goes

Every transfer has two prices. The first is the fee, quoted upfront. The second is the gap between the rate you are given and the mid-market rate, which is the real rate two currencies trade at between banks. A service can set the fee to nothing and take a percentage on the second price, and most customers will never see it happen, because nobody has the mid-market rate memorised.

This is why comparing fees produces the wrong ranking. An app charging a small flat fee at a good rate can be cheaper than one charging nothing at a poor one, on the same transfer. It is also why the honest services publish which one they are doing. Wise states that it uses the mid-market rate and charges a visible percentage on top. LemFi states that it takes an FX margin instead of a fee. Both are defensible. What is not defensible is the word 'free' with nothing after it.

Two men in London who worked this out first

Taavet Hinrikus was the first employee at Skype, living in London and paid in euros. Kristo Kaarmann worked in London, was paid in pounds, and had a mortgage in Estonia 🇪🇪 to service in euros. Each needed the other's currency, so from 2007 they simply swapped, month after month, using the mid-market rate they looked up online, and paid each other into local accounts. Neither of them ever moved money across a border.

What that year of swapping showed them was that their banks had not really been charging them a fee. The banks had been selling them a worse rate and calling the difference a service. They launched TransferWise on that observation in 2011, and the company, now Wise, built its entire pitch on showing the rate rather than hiding in it.

The turn is that the newest generation of Africa-corridor apps have gone the other way. They have taken the fee to zero, which is the part customers check, and they earn on the exchange rate, which is the part Hinrikus and Kaarmann founded a company to expose. That is not a scandal. It is a pricing choice, and it can genuinely be the cheapest option on a given corridor. It just has to be checked rather than assumed.

What the fifteen services actually do

The dataset is not a list of banks in the ordinary sense. Twelve of the fifteen are dedicated transfer services, two are neobanks and one is a card that sits on top of a UK 🇬🇧 current account. They solve different problems, and the choice mostly comes down to whether the person receiving the money has a bank account, a mobile wallet, or neither.

• · Mobile wallet delivery is where the diaspora-built services are strongest. LemFi, Taptap Send, Sendwave and NALA are all built around East and West African corridors first rather than reaching them as an afterthought.
• · Cash pickup still matters where the recipient is unbanked or rural, and that is Western Union's remaining argument: over 600,000 agent locations, at fees and rates that are consistently worse than the digital-first services.
• · Holding and spending in more than one currency is a different job again. Wise and Revolut do it, Currensea links a travel card to an existing UK 🇬🇧 account, and Majority offers a US 🇺🇸 checking account and debit card to people without a Social Security number.
• · European-origin corridors are their own market. TransferGo runs from the UK 🇬🇧 and 29 European countries to 164 destinations, which is a different map from an app built for Lagos and Accra.

The target nobody is close to

Sending money to Africa is the most expensive remittance business in the world, and it has been for as long as the price has been measured. The United Nations set a target for it in the Sustainable Development Goals: target 10.c commits countries to reducing remittance transaction costs to less than 3 per cent by 2030, and to eliminating corridors that cost more than 5 per cent. The World Bank publishes the running total every quarter in Remittance Prices Worldwide.

The global average is still above six per cent, and Sub-Saharan Africa is above eight, the highest of any region. On the $131.3bn of inflows in our money sent home dataset, a percentage point is real money, and it lands hardest on the countries that lean on it most. Comoros 🇰🇲, Gambia 🇬🇲 and Liberia 🇱🇷 each receive more than a fifth of their entire economy this way.

What to do with this

Pick two services that reach your corridor, put the same amount into both, and compare the figure the recipient sees before you send anything. Do it again in a few months, because rates and promotional pricing move and the app that won in March is not automatically the one that wins in August. If a service will not show you the recipient amount before you commit, that is the answer.