At Deskimo we operate in 13 countries, many of them across Asia, so our workspace partners speak many different languages. Underheard let us run qualitative interviews with them in their native language, and every conversation came back aggregated into one consolidated, actionable dashboard.
From those interviews we identified a set of concrete product features and prioritised them together with our partners, so we’re building what actually serves them best.
We weren’t actively looking for a tool like this when Underheard chose us for their pilot, and it has been extremely beneficial. We’ve since started rolling it out to Deskimo’s end users too, so we can collect their feedback the same way.
AI has read every word on the internet. We built a way for it to get what the internet has never had. Underheard is the AI interviewer that goes out and asks hundreds of people for you, then tells you what they think.
Want to try it on yourself first? Go to https://studio.underheard.ai/mic and tell our AI the most unhinged thing you've ever done. Then pass the mic to your friends and ask them anything. You don't even sign up, and nobody asks your name.
Imagine you're Red Bull designing your next crazy campaign, and you want to know what would actually impress people. Traditionally you'd interview thousands of them, which takes a couple of months. With Underheard it takes two days.
Or imagine you run a hotel and want to know what your guests thought of their stay, beyond the star rating. They tell Underheard about it, and it tells you what to fix. It works the same way for HR teams that want to know how a reorg really landed.
For decades, finding out what people think meant interviewers on the phone making one call after another during office hours. Good luck finding a German who'll do that on a Sunday. Underheard will.
You shouldn't need Red Bull's budget to find out what people think. Now asking a million people is as easy as asking one.
To ask your own audience, go to https://underheard.ai/start. Choose the setup that fits you best. Use our agent two weeks for free.
What would you ask the whole world? Tell me below. I'm Navid, and I'm here all day with Henri and Ben 🙏
Huge thanks to the 27 organizations that trusted us in our first three months ❤️
Love the idea, guys. If there’s one thing I’m absolutely horrendous at, it’s talking to customers. But I just want to say I tried the entire onboarding/demo, and I don’t think I’ve ever had a better time getting to know a product. Everything was super smooth, and having the AI assistant guide me through it is something I badly need for my own app, so I’ll be looking into that, because the value is just there.
Otherwise, the product itself looks really promising. My only worry is that people might be put off by the idea of being interviewed by an AI. I suspect if I don’t offer rewards, nobody’s really going to show up, but that’s on me to sort out the incentive. Have you seen decent completion rates without one? Either way, I’m not sure how great of a fit I am with a B2C but I think I might try it either way once I am ready to start a campaign.
You guys deserve to be higher on the list. Real gem, you should be proud.
Thank you @diae , this made our day! So glad the assistant made onboarding feel easy. That’s exactly what we were aiming for.
On the concern about AI, it’s a fair question. One of our first customers invited people to share feedback without offering a reward. Participation was similar to their usual feedback rounds, but the insights were much richer. There’s research on this too: in a study by Chicago Booth and Erasmus, 78% of job applicants given the choice picked an AI voice interviewer over a human.
We’ve also noticed that, for some topics, people feel more comfortable opening up to an AI. They know it won’t take offense, and it feels like they’re giving feedback to the company rather than to someone whose feelings they might hurt with negative feedback.
Incentives can help, but how you ask often matters more: keep it short, explain why their feedback matters (“we’re deciding X and would love your input”), and send the invitation through someone they already know.
Thanks again for the kind words 🙏
Thank you, Diae! Happy to hear the onboarding is good, we just added this recently as we hate onboarding usually on other apps and wanted to make it fun. Maybe we should really make this an extra service 😂
As for completion rate, we are with you. To be honest, it is not better than traditional surveys. Rather, we often have a traditional star rating as the first screen, with the interview being basically a qualitative part. So while slightly fewer respond, the information you get is a lot richer, which makes it way more complete. You can compare it between a quantitative study and qualitative research!
Again thanks for the warm words!!! And if you ever are interested let us know, happy to help a fellow out!
Can the interviews distinguish between people who say an idea sounds good and those who would actually pay for it? That’s the feedback I’d want most as an app developer.
Thanks for the great question!
Yes. Asking "would you pay for this?" mostly gets polite yeses, so we suggest starting with what people have already done: "Have you paid for anything to solve this?" "How much, and do you still use it?" The interviewer follows up on vague answers, so it can gauge how strong someone's intent to buy really is.
You don't have to write these questions yourself: the assistant in the studio can build the interview with you. Happy to help you get set up too!
Running market research used to mean weeks of scheduling interviews one by one. Now you describe what you want to know, and the AI holds the interviews for you and shows you what people actually think. If you have any questions, let us know!
Give it a try!
Get way better information from your customers in a few clicks.