Most SaaS founders ask the wrong questions. Learn how to run customer interviews that extract the insights that actually move the needle on retention and growth.

SaaS customer interviews: how to extract goldmine insights (updated 2026)

Most SaaS founders are running customer interviews wrong. They ask vague questions, get polite answers, and walk away thinking they've done user research. They haven't. They've done customer therapy.

The difference between a customer interview that extracts goldmine insights and one that wastes 45 minutes of everyone's time comes down to three things: who you talk to, what you ask, and how you shut up long enough to actually listen.

I've run over 200 customer interviews across Devlyn.ai, Laracopilot, and ViitorCloud in the last two years. Here's what I know: the answers that changed our product and growth strategy didn't come from the questions I planned. They came from the follow-up questions I asked when something sounded slightly off.

This guide breaks down the exact framework we use to run SaaS customer interviews that extract real insights. Not polite feedback. Not feature requests. The kind of insight that tells you why people actually stay, why they leave, and what they'd pay twice as much for.

If you want weekly tactics on SaaS growth from someone actively building, join my newsletter.

Why most SaaS customer interviews fail

The failure mode is predictable. A founder schedules a call with a happy customer, asks "what do you love about the product?", gets a feel-good response, and files it under "validation." Nothing changes. Nothing improves.

Here are the three reasons customer interviews produce nothing useful.

Talking to the wrong people

Most founders interview their happiest customers. That's a selection bias problem. Your happiest customers are not representative of the market. They're the ones who stuck around despite your product's rough edges because the core value was strong enough.

The people who reveal what's broken are the ones who left after week two, or who signed up and never activated, or who upgraded and then churned three months later. Those conversations are uncomfortable. They're also 10x more valuable.

When we were building Laracopilot, our first 50 interviews were with developers who had completed at least three projects on the platform. Great signal for understanding power user behavior. Zero signal for understanding why 70% of new signups never completed a single project. We needed to talk to the people who bounced, not the ones who stayed.

Asking leading questions

"Isn't it great that Laracopilot generates the Filament admin panel automatically?" is not a question. It's a statement with a question mark attached.

Leading questions confirm what you already believe. They don't reveal what the customer actually experiences. The customer says yes because they're polite, because they want to be helpful, or because they haven't thought about it enough to disagree.

Better: "Walk me through the last time you used the admin panel generation feature. What did you expect to happen, and what actually happened?"

Stopping at the surface

A customer says "onboarding was confusing." Most founders nod, take a note, and move on. The goldmine isn't that sentence. It's three layers deeper.

"What specifically was confusing?" "What did you try first?" "When you got stuck, what did you do next?" "What would have made that moment clearer?" That's where the insight lives.

The three customer segments that matter most

Before you book a single interview, decide who you're targeting. The answer depends on what question you're trying to answer.

Recently churned customers (within 30 days)

These interviews answer: why do customers leave?

Reach out within two weeks of churn. After 30 days, the memory fades and the story gets polished. You want raw, specific detail.

The offer that works: "I'm not trying to win you back. I genuinely want to understand what we got wrong so we can fix it for the next customer." That framing disarms them. You get the real story.

Newly activated customers (days 7-21)

These interviews answer: what's working about onboarding, and what's failing?

This window is critical because the experience is fresh. Ask them to walk through their first session from memory. The hesitations, the confusions, the moments where something clicked. This is where you find friction that your analytics can't show you.

Power users who expanded their usage

These interviews answer: what drove product-led expansion?

Find customers who started on a lower tier and upgraded without a sales touch. Understand exactly what moment triggered the upgrade. That moment is your growth lever. Replicate it.

The interview framework that extracts real insight

Before the call: do your homework

Spend 10 minutes reviewing their account data before you get on the call. When did they sign up? What features do they use? Have they contacted support? Have they upgraded or downgraded?

Walking into an interview without this context is like a doctor asking a patient to describe symptoms without looking at their chart. You'll ask questions you could have answered with three minutes of prep.

The opening that builds trust

Start with: "I'm going to be honest with you. I'm not here to pitch you anything. I want to understand your experience so we can build a better product. There are no wrong answers, and the most useful thing you can do is be blunt with me."

Then ask: "Can you tell me a bit about what you were trying to solve before you found us?"

That question opens the story. It establishes context. It positions them as the expert on their own problem, which they are.

The five questions that unlock everything

These are the five questions we ask in every interview, regardless of the customer segment.

1. "Walk me through how you first started using [feature]."

This generates a narrative, not an evaluation. Narratives are more accurate than assessments because they're grounded in specific events.

2. "What did you expect to happen, and what actually happened?"

This surfaces the gap between expectation and reality. That gap is where your biggest product improvements live.

3. "If you couldn't use our product tomorrow, what would you do instead?"

This reveals your real competitive set and how deeply embedded your product is. If they say "I'd just use Notion" when you think you're competing with Airtable, that's a signal.

4. "What's the one thing that, if we fixed it, would make this an obvious choice to recommend to someone else?"

This forces prioritization. You'll get one real answer, not a list of nice-to-haves.

5. "Tell me about the last time you were frustrated with the product."

Not "have you ever been frustrated." Everyone has. This assumes the frustration happened and invites them to describe a specific incident. Specific incidents give you specific fixes.

The follow-up that changes everything

The most important skill in a customer interview isn't knowing what to ask next. It's recognizing when to ask "tell me more about that."

When a customer says something vague, specific, or surprising, your instinct is to move on. Resist it. That vague thing, that specific detail, that surprising claim. Those are the moments to dig.

Marcus, a Laravel agency owner in the UK, told me in a Laracopilot interview: "The generation is fast, but I have to do a lot of cleanup after." I almost moved on. Instead I asked: "What does cleanup look like for you?" He described spending 40 minutes re-organizing file structure after every generated project. That led directly to a configuration feature we shipped three months later.

That feature is now one of our top-rated in post-session surveys. I almost missed it.

Want to go deeper on SaaS growth frameworks? Read my SaaS growth playbook for going from $0 to $1M ARR.

How to document and act on customer interview insights

Running great interviews and doing nothing with the data is almost worse than not running them at all. It creates the illusion of research without the benefit.

Build a jobs-to-be-done matrix

After each interview, fill in four cells:

Job: What were they trying to accomplish?
Struggle: What got in the way?
Trigger: What prompted them to seek a solution?
Outcome: What did success look like to them?

When you've run 10-15 interviews, patterns emerge across these four columns. Those patterns are your product roadmap.

Tag recurring phrases

Read through your interview transcripts and tag recurring language. When three different customers describe the same feature as "slow to respond," that's a performance issue. When five customers mention that they "couldn't figure out where to start," that's an onboarding problem.

The exact words customers use should appear in your marketing copy. This is called voice-of-customer copywriting, and it's one of the highest-impact things you can do with interview data. Your customers describe your product better than you do.

Create a "quote bank"

Every interview produces at least one sentence that captures something important. Pull these into a shared document. They're useful for:

Product team prioritization discussions ("here's what five customers said about this problem")
Marketing copy that resonates because it mirrors real customer language
Sales enablement: objection handling scripts built from real objections

A real example: how customer interviews changed our Devlyn.ai positioning

In early 2025, we ran 30 customer interviews for Devlyn.ai. The goal was to understand why customers chose us over cheaper offshore agencies.

We expected to hear "quality" and "communication." Those appeared. But the answer that appeared most consistently was something we hadn't anticipated: "risk."

Customers described choosing Devlyn.ai not primarily because of speed or quality, but because they felt safer. "I've been burned before" came up in 18 of 30 interviews. That's 60% of our customer base leading with a fear, not an aspiration.

That finding changed our entire positioning. We shifted from "AI-enabled senior developers who ship faster" to "the safe choice for founders who've been burned." Every piece of content, every case study, every sales conversation started addressing the fear directly.

Sarah, a fintech founder in Atlanta, said: "I'd spent $80,000 on a previous agency that delivered unusable code. When I found Devlyn.ai, what sold me was that you clearly articulated exactly what we'd get and what we wouldn't. No one had ever done that before."

That quote became the foundation of our homepage headline. We wouldn't have known to say it without the interview.

How often to run customer interviews

This is where most growth teams underinvest. Customer interviews aren't a quarterly ritual. They should be a monthly practice.

Here's a sustainable cadence that works without consuming your calendar:

Monthly: Two to three interviews with recently churned customers
Monthly: Two to three interviews with newly activated customers
Quarterly: Four to five interviews with power users or expanded accounts
Before any major feature launch: Five interviews with customers who've expressed the problem you're solving

That's roughly eight to 12 interviews per month. At 30-45 minutes each, that's four to nine hours. A high return on a modest investment.

The compounding effect is significant. After six months of consistent interviews, you develop a strong intuitive model of your customers' psychology. You stop being surprised by churn. You start anticipating what features will drive expansion. Your product decisions get faster and more accurate.

What to do before your next interview

Three things, in order.

First, pull a list of customers who churned in the last 30 days. Email them today. The script: "We noticed you cancelled. I'm not trying to change your mind. I'd love 20 minutes to understand what we could have done better. Would that be possible?"

Second, review the last three interviews you've done and identify the one follow-up question you didn't ask. The pattern of what you avoid reveals a lot about your blind spots.

Third, read everything you know about the customer before the next call. Account activity, support tickets, feature usage. Walk in knowing their story. You'll ask better questions.

SaaS customer interviews are the highest-value research tool available to founders who aren't yet at scale. They cost almost nothing. They reveal everything. You just have to run them right.

The bottom line on extracting goldmine insights

The founders who consistently build the right things aren't smarter than you. They talk to customers more often, they ask better follow-up questions, and they act on what they hear.

Running SaaS customer interviews that extract real insight isn't complicated. Talk to churned customers, ask "tell me more" one more time than feels necessary, and build a system for capturing what you learn.

The insights that change your product trajectory are sitting in the next conversation you haven't had yet.

I write about SaaS growth, customer research, and what's actually working at Devlyn.ai and Laracopilot. Subscribe to my newsletter for weekly tactics, no fluff.

Alpesh Nakrani is VP of Growth at Devlyn.ai and Laracopilot. He writes about building, growing, and scaling software companies at alpeshnakrani.com.