Build Your Career by Collapsing the Talent Stack

Belsky says small teams benefit when people can work across functions, tightening decision-making and reducing coordination overhead. At Adobe, groups of three to five tackle problems, while teams add more people and specialties when scale, geography, security, or enterprise sales require them.

Belsky says design and product management work best together, and sometimes one person can lead both. He distinguishes a product leader, who sets direction and keeps the team focused on its destination, from product managers, who coordinate execution and roadmap details. He also favors prototypes as a way to make product ideas tangible.

Belsky argues that a compelling vision can still miss product-market fit when founders do not understand customers’ real problems. He recommends observing customers at work and asking why they make specific choices, including needs for credit or reassurance. He also warns that products can become complicated when teams focus on vocal power users and neglect new customers.

Belsky says onboarding and the first-mile experience need ongoing attention because each new customer cohort has different expectations and tolerance for friction. He recommends progressive disclosure, shipping the minimum customers need, and polishing the features that distinguish the product. Teams should fix signup and time-to-value problems before launch, while leaving requests that arise from successful adoption for later.

Belsky describes Adobe’s approach of testing AI features in a standalone playground, grounding them in customer workflows, and reviewing product quality, integration, and ethics. For Generative Match, Adobe adds Content Credentials and retains the reference image to support transparency and respond to potential disputes. His four Ps for exploring AI are play, pilot, protect the pilot team with learning-focused measures, and invite provocation.