# The questions your customers are afraid to ask you

What permission to ask–and follow-up—looks like.

- Authors: Alex Tryon
- Published: 2026-11-06
- Category: Essay
- Canonical: https://www.askdewey.com/insights/the-questions-your-customers-are-afraid-to-ask-you/

---
This article was originally published on [*Fast Company*](https://www.fastcompany.com/91625640/the-questions-your-customers-are-afraid-to-ask-you)*.*

A few years ago, working as a fractional executive for a fast-growing startup, I picked up an unusual side gig: asking one “dumb question” at every all-hands.

We’ve all heard there are no dumb questions. We’ve also all learned, very quickly, never to be the one who asks them. This company had a version of that problem: All-hands open forums were met with silence. Onboarding sessions were full of new hires too afraid to put themselves out there. In the week before each, I’d listen to the office chatter and find the question I knew was living in at least a few people’s minds. If it was a little controversial, even better.

Asking these questions set the bar for what was acceptable. It gave the CEO an opening to address what needed addressing. It gave the room permission to follow.

That same dynamic plays out with your customers. The questions they most need answered are the ones they’ll never ask you unless you build the conditions, and the channel, that make asking possible. **The Filtered Version of Your Customer**

Companies invest heavily in understanding what their customers want. Surveys, Net Promoter Score, user interviews, and social listening. These tools are valuable, but they capture our most polished selves, not our raw ones.

Think about what each channel collects. Interviews capture what people are willing to attach their name to. Surveys record what customers believe will be helpful. Support tickets capture complaints, not curiosity.

Search was the first crack in that self-imposed filter. What people type into Google has always been [more honest](https://bookshop.org/p/books/everybody-lies-big-data-new-data-and-what-the-internet-can-tell-us-about-who-we-really-are-seth-stephens-davidowitz/9dc2d87dac42c7ba?gad_source=1&gad_campaignid=12440232635&gbraid=0AAAAACfld40jTHMeKZBcoJ4I3xOYtwrLa&gclid=Cj0KCQjwnbrUBhDOARIsAKKhPpfiDKsYmfXIvv0FziGZAeRDOOAGJn2lw4oyNghTbFnFoE6vGRgHLF4aAivkEALw_wcB) than what they told a human interviewer. But generative [artificial intelligence](https://www.fastcompany.com/section/artificial-intelligence) has blown the crack wide open. People now ask questions in [full sentences, with context](http://nber.org/papers/w34255), at the exact moment of need. These queries are longer, more specific, and more human than anything they type into a search bar.

Your customers are sharing their most honest, vulnerable questions right now, somewhere. By default, that somewhere isn’t you. It’s a generic large language model with no expertise in your domain and no obligation to tell you what it heard. It’s the richest customer data ever generated, and none of it is getting to you.

**What Unfiltered Questions Sound Like**

I’ve spent the last three years with an unusual window into that missing data. At Dewey Labs, we build search and Q&A tools grounded in expert content for public figures, associations, nonprofits, and brands. When their audiences ask the questions they’d never put their name to, we’re the ones who see them. We’ve processed more than a million questions, but it wasn’t the volume that changed how I think about customer insights. It was the sheer honesty of individual questions that struck a nerve.

I remember seeing “My kid is in love, what do I do?” in a client’s logs last year. As a parent, I recognized it immediately—vulnerable, urgent, and uncomfortable to ask with a name attached. No parent puts that in a survey. Few would say it out loud, even to a friend. Then, in another client’s logs, a different kind of vulnerability: “What does a judge do?” A completely valid question many of us could use a primer on, but one few would ask for fear of what it reveals about us.

That’s the quality unfiltered questions share. They show something the asker isn’t quite comfortable admitting, whether a fear, a knowledge gap, or a curiosity they’d be embarrassed to own. And context matters, too. Someone searching “Is melatonin safe for kids?” at 2:30 in the morning isn’t browsing. They’re desperate, and no keyword report captures that without the timestamp.

**Permission is How You Route the Traffic**

Every one of these questions is a business insight hiding in plain sight. The 2:30 a.m. timestamp tells you your users need you at hours you’re not available. The person asking “What does a judge do?” tells a civic newsroom it assumes too much. The question someone was too embarrassed to email you about tells you there’s demand you’ll never see in a feedback form.

But the insight only exists if the data comes to you. And that requires two things most organizations treat as separate projects.

The first is permission. Organizations that surface these questions publicly normalize the questions their audiences are hesitating to ask, whether in their content, their social presence, or their community spaces. For example, answering anonymously submitted questions on an Instagram Story opens a door for others to follow. And when a newsroom publishes an explainer on how voting actually works, it’s not dumbing things down. It’s sending a signal: You can ask that here. That signal is what pulls the private question out of the generic chatbot and into your world.

The second is a channel. Permission is necessary, but not sufficient. Nobody wants to ask, “Do I need to speak Italian in Italy?” to a person, but they feel free to ask a tool with no face and no capacity for judgment. This is what new tools like AI search or chat change for organizations. They provide a medium where your audience can reach you directly and where you can answer with your own expertise instead of leaving them to the internet’s average.

Do both, and something changes in the relationship itself. Companies that get this right don’t just have better data. They have a different relationship with their customers, one built on the implicit promise that no question is too embarrassing, too basic, or too raw to deserve a real answer.

**The Room is Already Asking**

At those all-hands meetings, the questions I asked out loud already existed. All I did was give them a safe place to land. And the company got to hear, for the first time, what the team was really thinking.

Your customers are in that room right now. The unfiltered questions are already being asked. The only thing left to decide is whether you’re the one who gets to hear them.