Stop Demoing. Start Listening Out Loud.

Most customer feedback calls are sales demos wearing a fake mustache.

We call them "roadmap reviews" or "vision walkthroughs." We say we want feedback. Then we share our screen for 40 minutes and leave five for questions. (Spoiler: there are no questions. Everyone has another meeting.)

I just had a 43-minute call that went the other way. By the end, the customer was designing features with us. He offered to set up 69 buildings himself and said yes to a pilot. Very little of that came from what I showed him.

So I went back through the transcript to figure out why. I found eight moves, and four things I'd still do better.

1. Huddle before they show up

The most useful 60 seconds of the call happened before the customer joined.

While we waited in the Zoom lobby, our CSM gave me three things. The attendees were from marketing. Reporting would land well. And they'd pushed the meeting back a month because they wanted to see something further along. Audience, hook and expectation, in under a minute.

I also told the CSM and our head of product marketing to jump in, add things, and stop me if I talked over anyone. Sounds small. It isn't.

Half the reason internal teams sit silent on customer calls is that nobody told them they're allowed to talk.

Try this: Before every customer call, get three answers from whoever owns the relationship. Who's in the room? What do they care about? What are they expecting to see?

2. Ask about their day before you show anything

Before I shared a single slide, I asked one question: where do you spend your time in our product, and where would you love to spend less?

He gave me gold. He runs a team of leasing agents. Every day he wants to know which applications are moving and which are stuck. If an agent had five apps at 85% yesterday, did any reach a decision today? To find out, he sets a filter and scrolls all the way to the right. Every. Single. Day.

That's a job to be done, described by the person doing it. No persona doc gets you that.

Then I said, out loud, that I'd use his answer to steer what I showed. That part matters. It tells the customer this isn't a scripted tour. Their answers change what happens next, so they give better answers.

Try this: Open with "Where do you spend your time, and where would you love to spend less?" Then tell them how you'll use the answer.

3. Tell the vision in their words

Yes, I still talked about the vision. But I built it out of what he'd just told me.

His "scroll all the way right" became my example of what we're fixing: information that exists but takes four or five clicks to reach. Our rule going forward is simple. Make the obvious thing actually obvious.

And since he lives in marketing, I explained our platform like a CMS. If we were WordPress, the thing our customers deliver to their customers is websites. For us, that front door is our AI leasing agent. Everything else sits behind it.

Nobody in marketing needs an architecture diagram. Everybody in marketing gets WordPress.

Try this: Pick one analogy from the customer's world before the call. Then swap at least one of your examples for one they gave you in the first five minutes.

4. Pause and test the why before the what

After the vision and before the demo, I stopped and asked: directionally, does this make sense?

That one pause pulled out the most important thing he said all call.

Instead of reacting to a feature, he told me about a real failure. Our AI agent books tours up to four hours out. His agents work Tuesday to Saturday, but the property owner wants leasing open Monday to Saturday. So on a Monday afternoon, the AI happily books a 5:30 tour... with nobody scheduled to show up. He's seen prospects sitting in their cars outside a building, and nobody knows they're there.

You don't get that story by asking "what do you think of this button?" You get it by asking whether the direction matches their reality.

Try this: Put a deliberate pause between the vision and the demo. Ask about direction, not features.

5. Treat the complaint as evidence

This is where most calls go sideways. A customer raises a problem, and we explain it away. It's a config issue. That's on the roadmap. Did you log a ticket?

I tried something else. First I told him he wasn't the first person to raise it. That flips the frame from "you're doing it wrong" to "you've spotted a pattern."

Then I named what I thought the real problem was. We deploy the AI as a black box. When it does something odd, nobody can see why, so it feels broken even when it's doing exactly what it was set up to do. And I asked him to tell me if I had that right.

He did. Now we weren't arguing about one bad booking. We'd agreed on a problem statement.

Then I jumped ahead in my demo to the diagnostics tool we're building, because it spoke to exactly that. (I literally said "I sort of sidetracked.") The script lost. The conversation won.

Try this: When you hear a complaint, name the underlying problem and ask the customer to confirm it. Then show whatever you have that speaks to it, finished or not.

6. Open the box

I was more candid about how our product works than most vendors would be.

I explained that we keep an internal health score for the AI: does it actually have the data it needs? I told him most of its odd answers trace back to gaps in the customer's own property data, not the AI. I walked through the two layers of setup that drive what it does. I even admitted the setting that fixes his phone-number problem already exists. It's just "buried in the spaghetti of our configuration."

That candor did two things. It built trust. And it made his feedback way more specific. Once he understood how it worked, he started telling us how he'd want it to work.

Same with trade-offs. He wanted direct control over those settings. I didn't promise it and I didn't refuse it. I said there's a real trade-off between control and risk, and I'd like to test it with him.

Try this: Explain one thing about how your product works that you'd normally keep internal. Name trade-offs out loud instead of promising your way past them.

7. Build together

By the second half, he was designing with us.

He took our data-quality idea and pushed it further: check the customer's own data, and audit every new property before it goes live. His example: "Your AI told us there's a pool. There's no pool." He suggested breaking the renter experience into stages, before the tour and after. And he pointed out that his agents send every application by hand after a tour. Why couldn't the AI send it and follow up?

I played those ideas back as I heard them. "So, taking this idea of data hygiene and bringing it forward to when a new property comes online." A playback proves you listened. It also lets them correct you on the spot.

I also said something vendors rarely say out loud: the owner of his portfolio has different needs than he does. They're at arm's length. He's on the ground. That told him his view counted, not just the view of whoever signs the contract.

Try this: Play back what you hear. Say whose needs you're designing for.

8. Invite, don't pitch

I closed with an invitation. When this gets closer, would you want to help us mock up your version of the workflow, maybe even pilot it? "I don't want to volunteer you for anything."

He said they'd love to.

An invitation gives them an easy way to say no. That's exactly why they say yes.

Try this: End with a specific, optional next step.

What I'd do differently

It wasn't perfect. Here's what I'd fix.

• I monologued. The vision section ran about six minutes without a break. It worked, but it was a gamble. Check in every couple of minutes.
• I made a timing promise on the spot. Good calls are contagious, and that's exactly when "hopefully by end of year" slips out. Great conversations are where roadmap commitments get made by accident.
• I got stories, not sizes. I never asked how often the no-show problem happens or what it costs. A great story with no number is hard to prioritize.
• I didn't recap, and my team barely spoke. No "here's what I heard, here's what happens next, here's when." And two smart colleagues mostly listened. Give everyone a job: one takes notes, one asks follow-ups, one owns the recap.

The cheat sheet

1. Huddle. Who's in the room, what they care about, what they expect.
2. Ground. Ask about their day before you show your slides.
3. Frame. Tell the vision in their words and their world.
4. Pause. Test the why before the what.
5. Follow the pain. Complaints are evidence. Let the demo follow them.
6. Open the box. Be candid about how it works and what it trades off.
7. Build together. Play back what you hear.
8. Invite. End with a specific, optional next step. Then recap it in writing.

Dear product people, next time you book a "feedback call," look at the agenda. If your screen share starts in minute two, it's a demo in a fake mustache.

Take it off. Ask about their day.