How to Find People for Customer Interviews in 2027

On this page

1. Decide who you actually need to talk to
2. Start with the people closest to you, carefully
3. Go where the problem is discussed
4. Write outreach that strangers answer
5. Screen before you book
6. Track recruiting like a small pipeline
7. Your next step

The interview questions usually get written first, often in one enthusiastic evening. That part feels like progress. Then comes the empty message window, the blinking cursor, and the slightly sick realization that you don’t know a single person to send it to.

A lot of interview plans stall right there. Laziness has little to do with it. Asking a stranger who has a specific problem for half an hour of their time feels awkward, and the first few attempts usually go nowhere.

This piece sticks to the unglamorous recruiting part: who to look for, where they hang out, what to write to them, how to screen, and how to keep a trickle of calls on your calendar. Starting something in 2027? All of it fits into this month.

Decide who you actually need to talk to

Write one line describing the person you want before you search anywhere. “Small business owners” fails that test. You can’t search for it, and you’d never know when you’d found one.

The lines that work describe a situation rather than a demographic, ideally one with a recent event attached:

• Freelance bookkeepers who switched software in the last year
• Parents who pulled a kid out of an after-school program this term
• Restaurant managers who still build the staff schedule themselves

Each gives you something concrete to ask about (“walk me through the week you switched”) and quietly hints at where those people gather.

Then add a disqualifier, meaning the person who looks like a fit and isn’t. Somebody who had the problem five years ago and solved it with a spreadsheet they love, for instance. Jot it next to your one-liner. The screener will need it later.

Start with the people closest to you, carefully

Your own network is fast and badly biased, because friends want to be kind to you and in an interview kindness is just noise. So skip interviewing them about the problem and ask them for introductions instead, with a note short enough to forward: “I’m researching how restaurant managers handle staff scheduling. Know anyone who does it? I’m not selling anything, I just want to hear how they manage today.” A warm intro from a friend of a friend tends to beat a stack of cold messages, and the person on the other end has no stake in flattering you. Once a call goes well, end it with one more question, “Who else should I talk to about this?” Anyone living with an annoying problem tends to know others in the same boat, and that one question can fill weeks of calendar on its own.

Go where the problem is discussed

People with your problem are often already complaining about it in public. Look in trade communities, groups organized around a specific job, the comments under how-to videos, local business associations and forums for particular roles. Lurk for a while first. What you want is a thread where someone describes the problem in their own words.

When you find one, reply to that person or message them directly:

I saw your post about juggling schedules across two locations. I’m trying to understand how managers deal with that. Would you be up for a 20-minute call?

It does better than a generic “looking for feedback” post because it answers something they actually said.

Groups have rules, and many ban self-promotion outright. Some want moderator approval for research requests too. Ask the moderator first; a lot of them say yes once they see you’re there to learn.

Offline counts too

Local customers, and people who work with their hands, are often easier to find in person. Shop owners have a quiet hour after the lunch rush. The supply counter at a trade store has a line of exactly the people you want. These chats run shorter than a booked call and tend to be blunter, which is a fair trade.

Write outreach that strangers answer

Most cold messages fail for two reasons. They’re about the sender, and they ask for too much. A message that gets answered fits on a phone screen and says why you picked this person, what you’re curious about, how long it’ll take and that no pitch is coming.

Hi Dana, I noticed you run scheduling for a two-location café. I’m researching how managers handle shift swaps and last-minute call-offs. Could I ask you about it for 15 to 20 minutes this week? No sales pitch, I’m still figuring out whether this problem is worth solving.

A thank-you helps, though it needn’t be big. Modest gift cards or a donation in their name are the usual choices, and with professionals an honest promise to share what you learned across all the interviews often lands just as well. Big incentives backfire. They pull in people who want the reward rather than people who have the problem.

Then send far more messages than feels reasonable (several times as many as the interviews you hope to book is a sane starting ratio). Lots will go unanswered. Silence from strangers says nothing about your idea.

Screen before you book

A wrong-fit interview costs two hours, yours and theirs. Two or three screening questions catch most of them.

Ask about recent behavior. “When did you last deal with a last-minute shift swap?” sorts people well. “Do you find scheduling frustrating?” doesn’t, since everyone says yes. Your disqualifier from earlier goes in here as well. The screener can live in a short form or simply in the first message thread, before any calendar link goes out.

And leave your idea out of it. Once someone knows what you’re building, they start telling you what they think you want to hear, and you lose the problem in their own words.

Neutral wording is easy to get from the free Customer Survey Generator, and its questions trim down into a quick screener form. The interview itself is a separate skill, covered in how to write customer discovery questions.

Track recruiting like a small pipeline

Say Marcus is exploring a tool that helps independent tutors deal with rescheduling. On Monday he asks three friends for intros to tutors they know, posts a question in two tutoring communities (after clearing it with the moderators), and messages five people who griped about cancellations in public threads. Four of those five never write back. The fifth, a math tutor who teaches out of her dining room, does, and at the end of the call she hands him two names. By Friday he’s had six conversations, half of them through referrals.

A week like that is about normal. Nothing in it is clever. Marcus kept a plain spreadsheet with the name, where he found each person, the date contacted, a follow-up date, status and who they referred, so nobody vanished into his inbox. He followed up once, a few days later, then let it go. He booked calls close together so patterns would jump out, with a gap after each one to write notes while they were fresh.

Five to ten good conversations usually show the main patterns. As soon as the calls start sounding alike, ease off. If each one still surprises you, keep going, or tighten your one-line description, because you may be talking to several different kinds of people at once.

Your next step

Today, write the one-liner and the disqualifier. Then send three intro requests to people you know and reply to two public posts where someone describes your problem. With luck, the first interviews land on your calendar before the week is out.

Curious which groups might react most strongly before you start recruiting? A simulated test can give you some hunches. Its participants are AI personas, not people you could ever call, so treat what it says as guesses to check in real conversations, never as a finding or a forecast. For how the two kinds of research fit together, read AI-simulated audiences versus real customer research. Once calls begin, the piece on spotting customer objections shows what to listen for.

Written by the SellTo.us Team. Examples in this article are hypothetical. SellTo.us results come from AI-simulated participants, not real people, and don’t predict sales or guarantee success.