Thin supply
The people who have already signed up to a US-based tester network are whoever found it, not a cross-section of how Kenyans pay, commute or shop. You can fill seats. You cannot assume they share a habit with your users.
Product research
UserTesting for Kenya: how to recruit Kenyan participants for product research
Global user-testing panels were built for markets where the supply already exists. Kenya is usually a thin slice of that supply: English-speaking, in Nairobi, comfortable on a laptop. If your product lives on M-Pesa, a mid-range Android and a Safaricom line, that slice will lie to you.
Product research in Kenya fails in a particular way. The session happens. The notes look tidy. Then the build ships, and the first real user cannot find the Paybill, loses the OTP when the screen sleeps, or never had the job you thought you were testing. The participant was not performing. They were the wrong person, recruited by a filter that treats “Kenya” as a country dropdown.
Afritesting is the recruitment layer for that problem. You keep Google Meet, Zoom, Teams or a Figma file. We match screened people, schedule them, and replace someone who does not fit the brief. You still run the research. We do not sit in the session for you, and we do not hand over the panel as a contact list.
Brief the behaviour, not the demographic.
“Kenyan adults, 18–45, smartphone” is how every panel looks the same. “Sent money to family outside Nairobi in the last 30 days, on the Android they already use” is a person your product will actually meet.
Where global panels slip
A Kenya filter is not a Kenyan participant.
UserTesting and the other international platforms are useful when the panel matches the market. For a lot of Kenyan product work, four things go wrong before the first task.
The screener’s English
An English-only screener selects for people who are fluent in screeners. The session is then full of habits translated into office English. Let people talk about the task in the language they use for it — English, Kiswahili or Sheng.
The wrong phone
Unmoderated tests get taken on whatever device qualifies the tester. Your users are often on a Tecno, Infinix or Samsung A-series, on a bundle, with one bar. That is a different product. Usability testing in Kenya is mostly this problem.
A freelance calendar
A dollar incentive and a laptop calendar assume a professional tester. Shop owners, riders and people on shift will not rearrange a day for that. Pay in KES, by M-Pesa or bank, for a session length that fits the work.
Method
Choose the session before you choose a platform.
Teams lose weeks looking for a “Kenya user testing tool”. The tool is not the hard part. The hard part is matching the question to a format you can actually field here.
01
Interview, if you need the why
Why they still keep cash after installing the app. Why the agent is safer than the screen. A video interview on your own Meet, Zoom or Teams link. Typical 45-minute consumer session: KES 3,500, incentive included.
02
Usability, if you need the break
Watch one flow fail. Moderated tests are the same live price as an interview. Unmoderated tests, once the build can survive without you, start at KES 2,000. Eight of those are about KES 16,000.
03
Survey, if you need how many
From KES 500 a completion. Fifty consumer responses are typically about KES 40,000. A survey will not show you which label failed. When you need patterns at scale, read survey participants in Kenya before you write 40 questions.
The brief
Recruit the last time they did the thing.
Recency beats identity. “Used a matatu in Nairobi this week” is a recruitment spec. “Interested in mobility” is a mailing list. Exclude people who can guess what you want to hear: staff in the category, colleagues’ friends, and anyone who has already sat three tests for you this quarter.
Screen for what they did, not for whether they are your target customer. A question that says “Do you send money upcountry at least twice a month?” teaches the answer. Ask which of several ordinary things they did in the last month, and bury the qualifying behaviour among them. More on how Kenyan samples get contaminated is in finding research participants.
What stays with your team
- The prototype, the live build, or the concept you want a reaction to.
- The task list, the consent to record, and what you do with the recording.
- The decision. Recruitment cannot tell you what to ship.
For a common consumer profile, the target is to start matching within 48 hours. A narrow profession takes longer. During the early service, no-shows and people who fail the agreed screener are replaced at no extra recruitment fee.
If the product moves money, do not stop at “uses mobile money”. Almost everyone you can reach does. The useful cut is the rest of the stack — bank app, Fuliza, SACCO, till, cash — which is the subject of fintech user research in Kenya.
Questions
Before you book a Kenya user test.
Is Afritesting a UserTesting alternative for Kenya?
Afritesting recruits and coordinates Kenyan participants for interviews, usability tests and surveys. You run the session in Google Meet, Zoom, Teams, Figma or whichever tool you already use. The gap it fills is finding people who match a Kenyan behaviour, not filling a seat labelled Kenya.
Can I user-test a Figma prototype with Kenyan participants?
Yes. Put the prototype link in the brief. For a first test in Kenya, a moderated session is usually safer than an unmoderated one, because you can tell when mobile data, language, or a missing payment step caused the failure.
How many participants do I need for user testing in Kenya?
Five to eight moderated sessions will show the breaks in a single flow. They will not tell you how the country feels about your brand. For that, recruit market research participants and, once the language is clear, run a survey.
Which Kenyan cities can you recruit user-test participants in?
Nairobi, Mombasa, Kisumu and elsewhere in Kenya. Name the places that matter. A Nairobi-only study is an honest sample. Calling it national is not.
Keep going
Adjacent problems.
Who should try the product?
Describe the behaviour, the place and the session. We’ll estimate the study before you pay, and you bring the tool you already use.