How Hint built the AI for the home on customer research from day zero

The company

About Hint

Hint is the AI for the home: an app that builds a living profile of your house from public records, environmental and utility data, and your own documents – and coaches you through home ownership. It launched nationwide on July 29, 2026, backed by $10 million in seed funding led by Slow Ventures.

Hint found its roots when co-founder Kyle Rush was building an AI to help homeowners navigate incentive programs. After buying a house of his own, he was introduced to the broader problem suite that homeowners navigate daily. That Easter, a friend asked what he was working on. He described the app he envisioned for coaching and empowering homeowners. "I've had this same vision for 40 years," his friend said. "I tried to build it once. Everyone told me it was too complicated." The friend was Martha Stewart. She soon joined Kyle and Yih-Han Ma as a fellow co-founder.

The challenge

A clear vision with open questions and no users – yet

The founders' vision was strong, but the usual questions for an early stage company prevailed: What to build first? Which features mattered most? How should the app feel when you use it?

Kyle knew it was important to make early bets carefully – and with real user data. He described his philosophy: "I know I'm going to be wrong. Solving any problem is going to take 100 guesses. I know the first 80 are going to be wrong. So let's just get the guesses out there as fast as possible and get signal on them."

But with no app and no audience yet, there was nothing to measure and no one to ask.

The solution

A feedback system to guide the product

The founders had experience with customer research in the past, but knew they were looking for a more modern solution this time around. A few key requirements guided their search:

Built for builders. The studies would need to be run by the people building the product, not a research department, and the whole team needs access to every response.

Audience included. With no users and no product at first, they required a solution that can curate a relevant audience.

Fast to launch, flexible for the future. They'd need to quickly launch studies while concepts were still rough, with the opportunity to probe deeper to understand the why behind users' feedback – and as their prototypes and questions evolved, they'd need those studies to be accessible and durable enough to inform decisions months later.

Voicepanel met all of Hint's needs. AI-moderated video interviews with probing follow-ups, first-pass synthesized findings, and CSV, API, and MCP access so studies start inside Hint's own tools. Voicepanel helped them find real homeowners from its consumer panels, screened to specific criteria, so they could run research before having a launched product.

Hint's customer-centric loop

Using Voicepanel, Hint built a product development loop centered around customers before they even launched the app.

Test guesses as early as possible. Value props, feature priorities, product personality, and the language of homeownership went in front of homeowners while still rough.

Make findings accessible and durable. Multiple humans review every study. The team writes its own synthesis with Claude's help, files everything in a shared library, and rolls up learnings monthly. The roll-ups feed a top 10 list of user truths, and every new hire and every project starts there. Entries include validated learnings like "homeowners want a coach, not a search engine," and "nobody wants a static maintenance list, they want advice for their own air conditioner."

Check decisions against research findings. A Claude skill points at the feedback archive, so as Kyle drafts specs and writes copy, he can check his ideas against real research.

Point the same method at experts. Realtors, home inspectors, builders, and former insurance adjusters answer Voicepanel studies on camera while the AI probes for detail. Their answers feed a proprietary expert database covering pool care, decks, and insurance – another source of Hint's differentiated first-party data.

The result

Launching confidently into the hands of every homeowner

Less than five months after the first research project, the app launched publicly with $10m in seed funding and coverage from TechCrunch, where Kyle confidently shared, "We want Hint to be in the hands of every homeowner in the country."

The clearest measure of the research is the product itself: Homeowners rejected automatic appointment booking, so scheduling shipped opt-in. They trusted recommendations only when well-cited, so recommendations show their source. Every meeting amongst the co-founders starts with what they're hearing from customers. There is no conversion lift or 10x improvement in this story – only a product that is ready for prime-time, driven by the voice of the customer.

With the app now publicly available, Kyle expects a resurgence of user testing on the same customer-centered loop that carried Hint to launch.