Between October 2025 and January 2026, we conducted 340 moderated usability sessions with first-time Bitcoin users in Kenya, Nigeria, Ghana, South Africa, and Ethiopia. Participants ranged from 18–55, were smartphone users, and had never completed a Bitcoin transaction before.
The task: receive 10,000 satoshis to a self-custody wallet. Simple in theory. The results were sobering.
"89% of participants could not complete their first Bitcoin transaction without abandoning or requiring intervention. The failure is not user failure — it is design failure."
The Top 5 Failure Points
Failure wasn't random. It concentrated in five specific moments, and across all five countries those same moments kept appearing. Fix these, and you dramatically close the gap.
Failure Point 1: The Bitcoin Address
A Bitcoin address looks like nothing users have ever sent money to. Phone numbers, bank account numbers, and Mobile Money IDs all have familiar patterns and predictable lengths. A 26–62 character string of mixed case letters and numbers does not.
What Happened
In 68% of sessions, participants expressed verbal uncertainty about whether they had the correct address before sending. Several typed addresses manually, introducing errors. The most common response was to simply stop and ask: "Is this right?" In a live transaction without a facilitator present, they would have abandoned.
The fear is rational. A wrong address means permanent, unrecoverable loss of funds. Yet wallets do almost nothing to help users verify they have the right one. Showing a name or label alongside the address — what we call address humanisation — reduced address-related abandonment by 34% in our Kenya case study. QR code scanning, which removes address entry entirely, lifted task completion from 31% to 78%.
Failure Point 2: Fee Shock at the Confirmation Screen
Not a single wallet we tested showed fee estimates before the final confirmation screen. Users would enter an amount, go through the full flow, and only discover the true cost — including network fees — at the very last step.
The Final Screen Shock
In Nigeria, 67% of participants either abandoned or reduced their transaction amount after seeing the fee for the first time on the confirmation screen. Across all five countries, 79% of participants were surprised by the fee shown at confirmation. In 34% of sessions, that surprise caused abandonment or a significant change to the transaction.
Users don't object to fees. They object to surprises. Showing an estimated fee range at the amount entry step — before users commit to the rest of the flow — reduced abandonment by 41% in our tests. This is a one-sprint fix that no wallet in our study had shipped.
Failure Point 3: Seed Phrase Backup
The 12 or 24-word seed phrase is the most important concept in self-custody Bitcoin. It's also the single best predictor of user abandonment. Across all five countries, 73% of first-time users either skipped the backup step entirely, made errors during it, or showed enough distress that they couldn't continue.
What Actually Happened at the Seed Phrase Screen
Skipped entirely (wallet allowed it): 41% of participants
Completed but with errors: 19% of participants
Completed correctly but visibly distressed: 28% of participants
Completed correctly and confidently: 12% of participants
Wallets present seed phrase backup as a technical ritual without explaining what it is, why it matters, or what happens if it's lost. Users told us they felt like they were doing something irreversible and dangerous — which they were — with no real understanding of the stakes. High stakes plus zero context reliably produces abandonment. Adding a plain-language explanation before the seed phrase screen reduced abandonment at this step by 30%.
Failure Point 4: No Feedback During Processing
After submitting a transaction, most wallets show a spinner or a vague "processing" message. On African mobile networks — where 2G and 3G connections are common — transactions can take 30–90 seconds to broadcast. That silence is unbearable.
"I pressed send. Nothing happened. I pressed it again. Then again. I didn't know if it worked or if I had sent my money three times."
Participant, Lagos, NigeriaIn 47% of sessions, participants tapped the send button more than once because there was no clear confirmation that their first tap had registered. Some submitted duplicate transactions. Others abandoned entirely, unsure whether their money had gone anywhere. Progress indicators and explicit state communication — "Transaction received. Confirming on the network…" — cut this failure mode almost entirely in the wallets that implemented it.
Failure Point 5: English-Only Interfaces
Even participants who rated their English proficiency as "good" consistently performed worse on English-only interfaces when the task involved financial decisions. Reading English and making financial decisions in English are different cognitive loads. Financial decisions require confidence, and confidence comes from your first language.
This was most severe in Ethiopia, where 91% of participants abandoned — driven almost entirely by the absence of an Amharic interface. But the language barrier created measurable friction in every country we tested. Participants in Ghana, Kenya, and Nigeria who encountered unfamiliar technical terms — "seed phrase," "private key," "wallet address" — frequently stopped and asked for clarification rather than proceeding.
The Five Fixes That Moved the Numbers
We worked with a Kenya-based wallet to implement targeted fixes for each of these failure points. Here's what happened to their abandonment rates:
| UX Change | Abandonment Before | Abandonment After | Change |
|---|---|---|---|
| QR code scanning promoted as default | 69% | 22% | -47% |
| Fee shown at amount entry | 61% | 36% | -41% |
| Seed phrase plain-language explanation added | 74% | 52% | -30% |
| Progress indicator added to send flow | 43% | 27% | -37% |
| Address humanisation (name shown alongside address) | 58% | 38% | -34% |
| Overall transaction abandonment | 67% | 22% | -67% |
Five targeted changes. No protocol modifications. No core architecture rebuild. Overall transaction abandonment dropped from 67% to 22%.
What This Means for Builders
None of these fixes require engineering heroics. They require someone to sit across a table from a real African user and watch them try to use your product. Most of the failures in this report would surface in a single day of moderated testing with five representative users. The problem isn't that builders don't care. It's that most have never watched a real user try their product.
1. Default to QR Code Scanning
Make QR scanning the primary address entry method, not an afterthought buried in the interface. Promote it visually. Task completion jumps from 31% to 78% when users scan instead of type.
2. Show Fees Before the Confirmation Screen
Display an estimated fee range at the amount entry step. This single change reduces abandonment by 35–41% across every African market we've tested. It costs one sprint to ship.
3. Make Seed Phrase Backup Human
Replace technical seed phrase screens with plain-language explanations. What is it? Why does it matter? What happens if it's lost? Use local analogies — a key, a PIN — not cryptographic terminology.
4. Communicate Transaction State Explicitly
Never show a silent spinner. Tell users exactly what is happening: "Transaction sent. Waiting for network confirmation." Estimate how long it will take. Confirm completion with clear, positive language.
5. Commit to One Local Language, Fully
Pick the most spoken language in your primary market and translate your core flows completely. A half-translated wallet is often worse than an English-only one — inconsistency erodes trust. Start with one language, do it properly, then expand.
The Bigger Picture
89% failure is not a user problem. The people in our research weren't confused because they were incapable. They were confused because the tools were built without them in mind. The 340 people in this research are traders, teachers, mothers, and farmers who came to a testing session because they wanted to understand Bitcoin. Most left having failed to complete basic tasks — not because they couldn't, but because the design failed them first.
The good news: the fixes are known, tested, and actionable. Every item in the table above is implementable in weeks, not months. If you're building a Bitcoin wallet for African users and you haven't run a moderated usability session in your target market, that's the first thing to fix. Everything else follows from watching real people use your product.
← Back to Home
