Yesterday
Ask what sold, what ran out, and what they did about it. An attendant remembers the rush. The owner remembers the gap. If the story is only “it was busy”, you have the counter.
Business owners
How to recruit SME owners in Kenya without interviewing the cashier
The person who can talk about invoices, stock and a failed Paybill is often not the person who answers WhatsApp. Consumer panels, event lists and “founders in Westlands” all miss the shop. The cashier will still take the seat if you let them.
SME research in Kenya falls over in two ways. You interview a nephew who was told to “represent the business”, and you design a dashboard nobody with a key to the shop will open. Or you interview the same ten people who attend every product breakfast, and you learn how startups talk about SMEs.
Recruit the decision. Who pays the supplier. Who notices that yesterday was short. Who deals with eTIMS, a till, or a book in a drawer. Afritesting screens for that and schedules around the shop. You do not get a database of businesses. You get sessions with people who matched. Because this audience is harder than a general consumer, the price sits above a standard KES 3,500 interview. The number is on the quote before you pay.
Do not ask for a KRA PIN to prove they are real.
A screener is not a compliance file. Someone who can narrate yesterday’s cash, the till or Paybill name, and who is owed money is the owner or the person who actually runs the books. A PIN number only proves they were willing to type one.
The cashier problem
Staff can take payment. They cannot close the day.
The rail
Till, Paybill, or cash in a box. They should know the name on the till and what happens to a payment that does not confirm. This is also why merchant fintech should not use a consumer screen — see fintech research.
The supplier
Who was paid last, in what form, and what is still owed. People who only serve customers cannot answer this without inventing it. You will hear the invention if you ask for a name and a weekday.
The books
A paper book, a WhatsApp thread, an Excel file, eTIMS, or nothing. You are not scoring their sophistication. You are checking they are the person who would have to change the system. Do not require a registered company if your users are not registered.
Logistics
Book the shop’s day, not your calendar.
No-shows here are often a delivery, a sick attendant, or a Friday that ran late. The fix is the slot, not a stern SMS.
Thirty minutes
One decision, not a tour of your roadmap. A consumer 45-minute block is a courtesy they cannot afford while the door is open. If you need longer, pay for longer and say so in the quote.
Off the peak
Hardware and wholesale have mornings. Salons and restaurants have a dead hour that is not yours to guess — ask. Do not stack five owners at lunch because your researcher is free.
Their floor, their phone
A Westlands boardroom full of people who can leave the shop is a different industry. Video is fine if they can step away. A prototype test has to be on the phone they use for the till, which is often not the newest one.
The half hour
Questions that earn the seat. Questions that waste it.
Waste: “Would you use a dashboard?” Every polite owner says yes. Waste: a feature list. Waste: asking them to pretend they are their customer. Buyers are a separate recruit — market research participants — and mixing them makes the owner perform.
Earn it:
- The last time a customer paid late, and what you did the next morning.
- How you know what to reorder, and the last time you were wrong.
- What you do when M-Pesa confirms on their phone and not on yours.
- Who else is allowed to see the day’s takings.
Nairobi is not the only place this works, and Westlands is not Nairobi’s trade. A shop on a busy street, a pharmacy in Kiambu, a hardware yard in Nakuru will teach you more about invoices than a founder who has never issued one. Say the counties. For common professional profiles, matching is aimed at a couple of days; a thin trade takes longer. If the person on the call is staff, they are the wrong participant — during the early service, mismatches are replaced without another recruitment fee.
If you are still building the screener itself, the general rules live in finding research participants. The extra rule for this audience is simple: if they cannot talk about money leaving the business, end the recruit and find the person who can.
Questions
Getting the owner in the chair.
How do you know a participant is the SME owner and not staff?
Ask them to describe yesterday’s sales, who pays the supplier, and whether customers pay a till, a Paybill, or cash. An attendant can take payment. They usually cannot say what was ordered, what is owed, or what happens when M-Pesa is down. Do not collect a KRA PIN in a screener.
How long should an interview with a Kenyan shop owner be?
Thirty minutes, booked off the peak — not Friday evening, not during a delivery. A 45-minute consumer slot is often too long for someone watching a counter. The incentive has to respect an hour they could have been selling. You see the full quote before you pay.
Why do SME studies cost more than a consumer interview?
Owners are fewer, busier, and easier to fake. A typical 45-minute consumer interview is KES 3,500. A narrower business audience is priced above that, because screening and replacements take more work. The estimate is shown before payment. There is no subscription.
Can I mix SME owners with their customers in one study?
No. Customers tell you why they buy. Owners tell you stock, credit, and what they do when a payment fails. Those are different decisions. Run them as separate studies, or the owner will perform for the customer and say nothing about the books.
Keep going
Related briefs.
Which owner, in which trade?
Tell us the business, what they must already do, and how long you can keep them. We’ll quote the study and find the person who runs it.