Launching a new product feature is always a leap of faith. Will the customer actually want it? Will it solve a real problem? And how do you know before you invest months building it? 

Now what if the feature involves art, one of the most subjective things in the world.

Those were the questions Noom faced when developing Mindful Art Break, an innovative feature designed to improve emotional well-being through guided art experiences. 

The challenge wasn't just building the feature, it was proving that it worked. How do you decide that something as subjective as art (and what type of art) can actually make people feel better?

Noom turned to research at scale. By gathering feedback from nearly 20,000 people using the SurveyMonkey platform, Noom transformed an emotional hypothesis into a feature backed by evidence.

On May 12, Noom’s product manager, Myles Johnson, joined SurveyMonkey’s senior customer success manager, Kalina Machlis, for a conversation about how his team transformed a bold idea into a feature backed by consumer data, and what every product team can learn from the process.

Watch the full session here.

Here are five lessons from Noom's research-first approach to product development.

1. Start small to learn fast

Johnson and his team ran small surveys testing different forms of art, exploring what resonated, what felt uplifting, what actually moved people. 

"A lot of the grunt work was done with just 100 people, 200 people, 500 people,” said Johnson, “It really allowed us to set a certain vision and direction for the project."

Those early surveys gave the team the confidence to refine the experience before scaling to larger studies.

Takeaway: You don't need thousands of respondents to shape your direction at the start, you need the right people, asked the right questions, early enough to actually change your mind.

2. Measure what matters – especially when it's subjective

Noom eventually scaled to a 5,000-person study. They presented 50 pieces of art in a survey, with 100 people evaluating and sharing their feelings before and after viewing each piece.

The goal was to understand how different people experience different types of art. Which pieces induced calm? Which pieces sparked joy? Which pieces helped people feel more connected, more focused, or more alive?

What they learned wasn't just interesting, it became the foundation for the final feature.

"To stay true to the feature, having humility and really being able to take a step back and say, okay, what is it that the people who are viewing this art need as opposed to what do we want this feature to look like—I think that was the difference between something that was successful and something that fell flat," Johnson explains.

Takeaway: The more subjective the question, the more valuable customer feedback becomes. Surveys make it possible to move beyond individual opinions and understand what resonates across hundreds of thousands of people.

3. Use AI to amplify human insight

With 5,000 responses from people viewing 50 pieces of art, Noom had something precious: a map of human emotion.

They used that map to build a proprietary AI algorithm that scored 20,000 pieces of public-domain art. The algorithm didn't crown a "best" painting. Instead, it understood something far more useful: what emotions resonate when people view specific work.

This is where AI gets interesting, not as a replacement for human judgment, but as a translator of it.

The person curating art for the feature each week still makes creative choices. But now they make them with precision. They understand not just what art looks good, but what feelings it catalyzes. They can pair that knowledge with the algorithm's insights and say: "This piece from this era will resonate with people seeking calm. That piece—which I didn't expect—also induces that same calm."

Creativity doesn't disappear when you add rigor. It sharpens. It gets smarter.

"AI is very malleable," Johnson notes. "What you feed it really can determine the direction that it can go." In Noom's case, what they fed it was human truth. The algorithm learned to serve that truth, not replace it.

Takeaway: AI is only as valuable as the insights behind it. Research came first. AI made those insights scalable. 

4. Research helps you move faster

Johnson's advice is blunt: "Market research is a humble skill."

Research requires admitting your hunch on what your customers want might not be right. It requires getting feedback from real people, and making decisions based on that feedback. , you. 

The teams that skip this step usually justify it with speed. "Research slows us down," they say.

Noom discovered the opposite.

Those early studies with just a few hundred participants helped the team move faster by pointing them in the right direction. The larger study then validated those insights and helped refine the final product.

They didn't trade speed for rigor. They used rigor to move faster.

The result: hundreds of thousands of users, external recognition, a Webby Award nomination, and, most importantly, a feature that helps people feel better, which was the goal all along.

Takeaway: Research isn't a detour. Done early, it's one of the fastest ways to reduce risk and build confidence before launch.

5. Let curiosity lead the product

What's striking about Noom's approach is how it flipped the traditional product launch script.

Most teams ask: How do we build this? Then: How do we convince people to use it?

Noom asked first: What do people actually need? Then: How do we build it to deliver that?

It's a small difference in framing. It makes an enormous difference in outcomes.

"Curiosity was running ahead of everything else," Johnson says about the feature's development. "It made it very easy to be a part of and be invested in and be interested in figuring out how we were going to do this the right way for our users."

That curiosity wasn't a nice-to-have. It was the engine. When you start by genuinely wanting to understand people—not convince them—everything that follows is easier. 

Takeaway: The best product teams don't start with answers. They start with better questions.

Put this into practice

Noom's success wasn't just the result of a great idea. It was the result of asking the right questions, and listening.

If you're bringing a new idea to market, SurveyMonkey LaunchPad gives you the tools to validate concepts, test messaging, measure purchase intent, and gather customer feedback before you launch, so you can move forward with confidence, not assumptions.

Want to watch more conversations on how businesses are turning curiosity into better decisions? Check them out here CuriosityCon.