Will Wu on Human-Centered Product Design, Snap and AI

Will Wu, Match Group’s CTO and a former Snap product leader, discusses human-centered design, team culture and the relationship between creative intuition and evidence. He draws on work including Snapchat Discover, Chat and Snap Games.

Wu grew up around science and technology: his father worked in semiconductors and his mother was a chemist. He recalls taking computers apart and learning from people he met on Internet Relay Chat, where curiosity about networking and security became a formative experience.

After studying engineering and beginning a messaging startup, he met Snapchat co-founder Evan Spiegel in 2013. A conversation about their shared enthusiasm for the original iPhone led to a product-designer offer. Wu says Spiegel helped him see himself as a product person rather than only an engineer.

Wu credits Spiegel with strong intuition about social interaction, design and communication. He considers the ability to explain a product vision an important part of product leadership. His praise reflects a close former working relationship.

The team spent more than a year building Discover, but its initial placement made the feature hard to find. Moving it one screen closer to friend stories changed its visibility, Wu says. The lesson was to understand existing traffic patterns and place a new feature where users can naturally encounter it.

Snap Games gave him a different set of lessons. Wu had not previously managed a team or acquired companies, and the launch required coordinating developers across countries while preparing for public presentations and media questions.

The stressful experience gave him confidence that he could handle challenges beyond his prior experience. He says he still feels impostor syndrome but increasingly treats the discomfort as a sign that he is learning, rather than a reason to stop.

Wu sees product as both art and science. At Match Group he was beginning to build a team called the Art and Science Lab, bringing creative and technical perspectives together. In his view, intuition is especially important when an entirely new product has little data, while evidence helps guide iteration.

Human-centered design means keeping the people who use a product in mind from early need-finding through prototyping, launch and revision. Wu warns that a metric can point toward a change that looks successful locally but harms the user experience over time.

He uses Apple Watch features such as fall detection and a hand gesture as examples of technology designed around human needs. He contrasts this with his frustrating experience using enterprise expense software, while acknowledging that the enterprise buyer and the end user can have different priorities.

For an early founder, Wu recommends talking to target users directly and listening for their problems. In online dating, for example, creating a profile can feel awkward and choosing among apps can be difficult. He suggests documenting those needs before bringing people from different disciplines together to explore solutions.

Simplicity is not always better, he says, but it is particularly important for a novel product. A simple core can help people understand and adopt it; after product-market fit, a team can add sophistication without making the initial experience inaccessible.

To avoid feature creep, Wu thinks about information architecture: common actions should remain approachable while deeper settings can be available to power users. He arrives at that balance through repeated designs, prototypes and user feedback.

He likes rapid prototypes built in working software as well as interactive mockups. They allow a team to feel an idea, test it informally and iterate before release. For a new launch, he wants a product the team is proud of, while recognizing that waiting for perfection can delay learning about a real need.

For creative brainstorming, Wu prefers a small in-person group, a relaxed setting and psychological safety. He tries not to dismiss unusual ideas too quickly: even a half-joking suggestion can contain a valuable insight.

His rough preference is no more than about five people with different skills and mutual trust. An agenda can help keep a discussion useful, but excessive structure can suppress the creative process. He seeks a balance rather than a fixed rule.

Diverse perspectives and direct observation of users help generate ideas. Wu suggests learning from disciplines beyond software and seeing the target customer’s environment in person, so the team can understand the problem with more nuance.

Revenue and innovation both matter. Wu has found it useful to give each goal a dedicated team, because a new experiment can temporarily conflict with current revenue targets. The teams still need communication, trust and a product leader who balances their work.

He wants a culture of humility, empathy and learning. In his experience, an employee who consistently dominates discussions or dismisses others can discourage the team. His view of culture fit is a management judgment and needs to be applied fairly.

When hiring product people, Wu looks closely at work they have built or imagined. He asks why a particular interface element or piece of copy exists, using the answer to understand the candidate’s thinking and interest in the craft.

While building Snap Games, he found an Australian studio through a multiplayer game whose visual design and interactions matched his vision. He says the work convinced him of the team’s abilities before he met them; an introductory conversation then helped him assess whether they could work together.

He looks for curiosity, adaptability and a desire to learn in designers, engineers and product managers. Asking about a candidate’s long-term ambitions helps him understand motivation and whether he can support their growth.

Wu values passion and work-life balance rather than assuming everyone needs to be in an office every day. He worries that some people may pursue product management solely because it is fashionable or well paid; that is his hiring preference, not a judgment about everyone’s motives.

He does not personally rely heavily on artificial case studies or brainteasers. Reviewing past work and asking about concrete choices gives him a better signal in his experience, although he acknowledges such tests can measure some forms of thinking.

One of his own hiring mistakes is giving people too long when signs suggest a poor fit. Wu says his optimism can make him overlook feedback from the team. Stebbings asks whether being motivated by money is necessarily wrong; Wu says he prefers deep enthusiasm for the product’s impact.

He does not require every candidate to have worked in the same industry. A different background can bring useful first-principles questions; he points to the game-development community as a source of strong product thinking outside games.

Wu says designers on his team were already experimenting with generative image tools for concepts. He expects multimodal AI could eventually provide quick feedback on mockups from simulated user perspectives. This is a prediction, and simulated feedback is not evidence that real users will respond the same way.

He disagrees that AI will make user interfaces irrelevant. Conversational interfaces open new ways to interact, he says, but command lines, graphical interfaces and touchscreens have continued to coexist. He expects screen-based interfaces to remain useful alongside conversation.

Wu is enthusiastic about generative AI while warning that an application built only on prompts to a widely available model may lack durable differentiation. His broader claims about model capabilities and future costs are opinions expressed at the time, not guarantees.

At Match Group, he views a user leaving because they found a fulfilling relationship as a success, even though it means losing that user. The business must also reach new generations. He acknowledges that dating apps compete for limited attention with games and other mobile products; he has a commercial interest in Match Group.

Wu says teams should normally listen to users even when a request conflicts with a core product value. The task is to understand the underlying need and look for a solution consistent with the product’s principles.

His advice to a new product leader is to keep building and practicing, including personal projects, and to treat the discomfort of learning as an opportunity. He also says he would act earlier on persistent behavior that undermines a team’s culture.

Outside work he enjoys watches as a blend of art, technology and personal expression. He likes the way an object can commemorate a milestone and acquire meaning over time.

Wu highlights a Tinder feature intended to help users select profile photos using AI. As Match Group’s CTO, he has a direct interest in the product; its privacy and effectiveness are described as company claims and his hopes, not independently verified results.

He hopes that reducing the effort of choosing photos will remove friction for people starting a profile. Stebbings thanks him for the conversation, and Wu closes by thanking the host.