Why Amazon and Chobani Use AI Interviews to Get Brutally Honest Customer Feedback

Akram Chauhan
Akram Chauhan
7 min read149 views
Why Amazon and Chobani Use AI Interviews to Get Brutally Honest Customer Feedback

Picture this: a product designer spends weeks perfecting a new feature. They finally get on a Zoom call with a customer, hold up their creation, and ask, "So... what do you think?" The customer, not wanting to be rude, smiles and says, "It looks great!" But in reality, they're confused, frustrated, and would never use it.

This little white lie is the bane of product teams everywhere. Traditional customer research is not only painfully slow—often taking up to eight weeks—but it's also plagued by the human desire to be agreeable. We don't want to hurt anyone's feelings, so we soften our feedback. The result? Companies build products based on polite fiction instead of hard truths.

A fast-growing startup called Strella is flipping this dynamic on its head, and it's catching the attention of giants like Amazon, Duolingo, and Chobani. They've built an AI that conducts customer interviews, and it turns out people are far more willing to share their unfiltered, brutally honest opinions with a machine. Fresh off a $14 million Series A funding round, Strella is proving that the key to understanding your customer might just be to take the human interviewer out of the equation.

The Uncomfortable Truth: Why We Lie to Humans but Tell AIs Everything

One of the most fascinating discoveries to come out of Strella’s platform is a simple quirk of human psychology: we're more honest with AI.

Lydia Hylton, Strella's CEO, saw this pattern emerge again and again as clients ran A/B tests comparing human-led interviews with their AI-moderated ones. "If you're a designer and you get on a Zoom call with a customer and you say, 'Do you like my design?' they're always gonna say yes," she explains. "They don't want to hurt your feelings."

With Strella's AI, that problem vanishes. Participants don't feel the social pressure to be nice. They just speak their minds. "They would tell you exactly what they think about it, which is really valuable," Hylton adds.

This isn't just a theory; it's driving real business results. Brian Santiago, a Senior Product Design Manager at Apollo GraphQL, put it plainly: "Because participants open up more with the AI moderator, the feedback is deeper and more honest."

Initially, companies were skeptical. Priya Krishnan, Strella's COO, notes that clients worried about "eroding quality" by using an AI. But the reality was the exact opposite. "The level of insight that you get is much richer because people are giving their unfiltered feedback," she says.

Compressing an 8-Week Marathon into a 48-Hour Sprint

Beyond honesty, the biggest pain point in customer research has always been speed. The traditional process is a soul-crushing marathon:

  1. Write detailed interview guides.
  2. Recruit and screen participants.
  3. Schedule dozens of calls across time zones.
  4. Conduct the interviews (and take frantic notes).
  5. Transcribe hours of audio.
  6. Synthesize mountains of qualitative data into key themes.
  7. Create a presentation to share the findings.

This entire ordeal can easily eat up two months of a highly-skilled team's time, delaying critical decisions and slowing innovation to a crawl.

Strella automates what Hylton calls "the middle 90% of the work." The platform uses an AI to moderate free-flowing, voice-based interviews that feel like a natural conversation. The AI asks questions, probes for deeper insights on interesting responses, and then automatically synthesizes the findings. It generates highlight reels, charts, and summaries from hours of unstructured chatter.

The upshot? "It used to take eight weeks. Now you can do it in the span of a couple days," Hylton says. The company claims an average time savings of 90% on manual research tasks. This frees up researchers, designers, and product managers to focus on the two things that actually matter: figuring out the right questions to ask and deciding what to do with the answers.

Finally, a Real Solution for Mobile App Research

If you've ever tried to understand why users struggle with a mobile app, you know it's a massive headache. You can't just ask them; you need to see what they're doing. But getting a user to share their phone screen reliably during a research call has been a technical nightmare.

This is where Strella is building a serious competitive advantage. A major focus of their new funding is to expand their mobile application, which enables persistent screen sharing throughout the AI interview.

"We are the only player in the market that supports screen sharing on mobile," Hylton states. Now, a company like Duolingo can watch a user navigate their app in real-time while the AI asks questions like, "What were you thinking when you tapped that button?" or "What are you looking for on this screen?" It's a game-changer for understanding mobile user experience, and for some of Strella's customers, it's the first time they've been able to conduct this kind of research at scale.

This approach is also incredibly engaging for participants. While a typical 60-minute survey sees drop-off rates of 60-70%, Strella's conversational interviews of the same length boast nearly 100% completion rates. People are simply more willing to talk than to click through endless radio buttons.

How Strella Sidestepped the Hype to Build Something Real

In a market flooded with AI startups, Strella's founders made a crucial decision that set them apart. Their initial idea was to build "synthetic respondents"—AI-powered digital twins that could simulate customer feedback. It was a trendy concept, but they quickly discovered a fatal flaw.

"We actually pivoted from that," Hylton admits. "People are very intrigued by that concept, but found in practice, no willingness to pay right now." They realized that while you can just ask ChatGPT what a user might think, it's no substitute for talking to a real person.

So, they focused on a harder problem: collecting proprietary qualitative data from real humans, at scale. Their bet is that the true, defensible value isn't in simulating customers, but in building a rich, searchable "system of truth for all qualitative insights" within a company, as Lindsey Li of Bessemer Venture Partners puts it.

Their technology is also fundamentally different from competitors who often started with text-based surveys and later bolted on clunky voice features. Strella was built from the ground up for free-form conversation. There's no typing, no buttons, no waiting for the next question to load. It's just you and an AI, talking. That seamless experience, Hylton says, is "an extraordinarily hard product to build" and represents a significant technical moat.

Not Just Stealing Budget, But Creating It

Perhaps the most telling sign of Strella's success is that they aren't just winning over companies with existing research departments. They're creating entirely new markets.

"Several of our customers didn't do research before," Krishnan explains. They always wanted to, but lacked the dedicated staff or resources. "They have purchased Strella to kick off and enable their research practice."

This is happening at a critical time. Forrester's 2024 Customer Experience Index found that CX quality has declined for three straight years—an unprecedented trend. Companies are desperate for better ways to connect with their customers, and Strella is making it accessible.

The platform's mission is to "democratize access to the customer." Instead of waiting weeks for a formal report, anyone in the organization can get answers. Imagine an engineer who is skeptical about a proposed feature change. With Strella, a product manager can instantly pull up a video highlight reel.

"You can say, 'Okay, engineering team, we need to build this feature. And here's the customer actually saying it,'" Hylton illustrates. "'This isn't me. This isn't politics. Here are seven customers saying they can't find the Checkout button.'" That direct, visual evidence is far more powerful than any PowerPoint slide.

The Future is a Human-AI Research Team

With $14 million in fresh capital, Strella is focused on scaling. They've found their product-market fit—proven by 10x revenue growth, zero customer churn, and a 100% conversion rate from pilot programs to paid contracts. Now, it's about execution.

The product roadmap hints at an even more integrated future. They're working on adding computer vision, allowing the AI moderator to react to facial expressions and non-verbal cues. They're also building more sophisticated collaboration tools. Soon, a human researcher might listen in on an AI-led call and jump in with a specific follow-up question, or have the AI act as a backup, taking notes and suggesting probes while they lead the conversation.

This vision of human-AI collaboration is key. Strella isn't trying to replace the human researcher. They're trying to superpower them by automating the tedious parts of the job. The strategic work—the empathy, the curiosity, the critical thinking—remains human.

By collecting and organizing real customer conversations at scale, Strella is building more than just an interview tool. They're creating a living, breathing archive of customer truth, one that anyone in a company can access to make better, faster, and more empathetic decisions. And that's a conversation worth having.

Tags

AI Startups Customer Research User Feedback Enterprise AI

Stay Updated

Get the latest articles and insights delivered straight to your inbox.

We respect your privacy. Unsubscribe at any time.

Aicosoft

AI & Technology News, Insights & Innovation

AICOSOFT delivers cutting-edge AI news, technology breakthroughs, and innovation insights. Stay informed about artificial intelligence, machine learning, robotics, and the latest tech trends shaping tomorrow.

Connect With Us

© 2026 Aicosoft. All rights reserved.