why prototype first

• Not about making it pretty - it's about not building the wrong thing

• AI can write code fast, but it can't tell you if your users will hate it

• Moving pixels is cheaper than moving features

• Each early pivot costs you hours instead of weeks

• Your engineers should build solutions, not figure out problems

when you can skip prototyping

• Internal tools where you know exactly what you need

• Super technical products where UX isn't the main thing

• When your team has built something very similar before

• Small features in existing products

• When speed > perfect solution

when to absolutely prototype

• New market you don't fully get

• Complex user flows

• Anything involving money/payments

• Features that can't fail

• When multiple stakeholders need to align

• Products where UX is make-or-break

ergo: hybrid

• Use AI tools for quick code experiments

• Prototype only the risky parts

• Build simple stuff directly

• Design complex flows first

• Let engineers challenge your assumptions early