n.o
LinkedInInstagram
Case Study

myMTN App — Accessibility for Visually Impaired Users

RoleUX/UI Manager
PlatformMobile App (iOS & Android) + Voice
Timeline2024 – Present

Background

Accessibility is often evaluated by proxy — an automated scanner, a checklist, or a sighted designer navigating with their eyes closed. None of that tells you whether a visually impaired person can actually open the myMTN app and send money to someone on their own.

In 2024, the Digital Channels team set out to answer that directly. Rather than auditing the app against a standard, we brought real visually impaired users into the room, gave them their own devices, and watched — alongside developers and IT — where the experience broke down.

The
Problem

The myMTN app had never been evaluated end-to-end by the users least likely to be well served by it. The main issues:

  • No verification that visually impaired users could independently complete core journeys — sign-up, login, buy airtime, buy data, send money — using TalkBack (Android) or VoiceOver (iOS).
  • Screen reader behaviour was inconsistent and, in places, actively broken: contact fields were not announced, dropdowns were not indicated, "MTN" was mispronounced as "Mountain," phone numbers were read aloud as a single indistinct figure instead of individual digits.
  • No structured process existed to catch these issues before release, or to prioritise fixing them systematically afterward.

Specific goals

Establish whether visually impaired users can navigate, interact with, and derive value from the app, and close the gap where they can't.

Solutions - Phase 1

The Solutions - Shipped & In Progress

1

Android — largely resolved

Twelve issues logged; ten shipped. Fixes included:

  • Contact picker bug not read as a contact picker entry point,
  • Unindicated dropdowns now working
  • The "MTN"/"Mountain" mispronunciation resolved
  • Illegible phone-number readout resolved. Phone numbers are no longer read as one number but as a single digits for a phone number
  • Sections that had missing descriptions are now being read
  • Button names being read with underscore in the name e.g. button_login_bt, resolved.
  • Missing alt voice description for Homepage imagery resolved
2

iOS — in progress

  • Alt text for onboarding and login
  • Structured headings
  • Alternative voice description for Homepage imagery

Research &
Discovery

How we discovered the issues

In-person Usability Testing

We invited a number of our visually impaired users for an in-person usability test. We had to see the problems for ourselves.

We gave them prompts on the key journeys users usually perform on the app. The journeys were:

  • Onboarding & Login
  • Airtime purchase
  • Data purchase
  • Send MoMo (Mobile Money transfer)

We connected users' phones to large screens so the whole room could observe their interactions on the app.

It was amazing to see that visually impaired users hold their phones to their ears and not to their faces when performing an action on it. It was insightful.

Shots From The Session

Usability testing session: participant using the myMTN app while the room observes on a large screen
Usability testing session: team observing a visually impaired participant navigating the app
Usability testing session: participants and stakeholders around the testing table
Usability testing session: wide view of the room during the accessibility test

Solution - Phase 2

Voice AI is coming.....

The next step is to integrate a speech AI solution into the app, where users can speak to the app to take actions, mainly to buy airtime, data, send MoMo (send money), and check their balance. This is still a work in progress.

Measurable success

Impact

19+ accessibility barriers flagged19+ distinct accessibility barriers identified across sign-up, login, airtime, data, and money-transfer journeys
10 of 12 Android issues resolved10 of 12 logged Android issues resolved, with the remaining 2 in backlog
Accessibility testing converted from a one-time project into a standing practice, now embedded in both engineering QA and research recruitment
Findings directly shaped the myMTN redesign roadmap and a prioritised remediation plan for Abena AI's voice experience