Mock Polishing

Mock polishing is the detailed refinement process where senior product designers transform rough junior mocks into polished, high fidelity designs ready for handoff. Seniors would adjust every element to match the design system exactly. They fixed broken auto layout constraints. They applied the correct design tokens for colors, typography, spacing, and elevation. They perfected visual hierarchy so the important elements drew the eye first. They designed all the component states including hover, active, disabled, error, and loading. This work happened in Figma at companies like Google in 2023 and Meta in 2024. It was how seniors demonstrated their value and protected their high salaries. The task required deep knowledge of the brand and system that took years to build. Many seniors spent the majority of their week in this mode because it felt like real craft. The output looked beautiful in presentations and case studies. It created the illusion of progress while the actual product often looked different in production.

It is not strategic design work or solving ambiguous problems. It is not the initial brainstorming or wireframing of new features from user research. Those tasks involve creativity and judgment in uncertainty. Mock polishing assumes the hard choices are made. The senior just executes on taste and consistency. It is not building the actual product either. The polished mock served as a proxy for the real interface. Engineers then interpreted it as best they could. This created a game of telephone that AI bypassed completely by generating code directly. Mock polishing also is not a moat anymore. When the work could be described in a prompt the scarcity disappeared. Taste still matters but the delivery mechanism changed from manual refinement to AI direction and selection.

Here is a concrete example from the 2024 workflow at a company similar to Dropbox. A junior designer submits frames for a new sync feature. The screens have decent layout but the details are off. Padding is inconsistent. Type styles do not match the library. Icons are wrong. The senior designer opens the file and goes to work. Over the next two days they realign everything to the eight point grid. They swap in design tokens from the central library. They add shadows with the correct blur and opacity. They design variants for mobile and desktop. They create a full set of states for the new toggle component. The final file gets presented in critique where everyone praises the polish. The feature launches two weeks later after handoff. Now replace that with the 2026 workflow at Linear. The same feature request comes in as a ticket. The design engineer pastes the description into Cursor along with the design system context. Claude generates the React code using Tailwind classes that match the brand perfectly. The engineer adjusts for perfect taste in key areas. They push a PR that same afternoon. The feature ships the next morning. No Figma polishing occurred. The live product is the only artifact that matters. A second example comes from Anthropic in 2025. They needed to update the settings panel in claude.ai. Instead of polishing mocks the team used v0 to generate the new surface from a text prompt that included prior critiques. The design engineer selected the best output, wired up the interactivity in code, and deployed it. User feedback came directly on the shipped work. The speed increase was massive. What used to be a two week design cycle became a two day cycle. One more example from a freelance senior who worked with multiple SaaS companies in 2024. Each client expected two rounds of polishing on every flow. The senior billed forty hours a week just on refinement. In 2025 that same freelancer retrained, bought a Cursor subscription, and now delivers complete shipped features in the time it used to take to polish one flow. Their rate increased thirty five percent.

Use mock polishing when your organization has not yet adopted AI generation tools and still values perfect Figma files as the primary deliverable. This remains true at some enterprise companies with heavy compliance needs or slow decision making cultures. It makes sense for brand heavy marketing websites where every detail contributes to perception. Use it to mentor juniors on the subtleties of good design. Use it in 2022 at Airbnb where the design bar was exceptionally high and consistency was non negotiable. Never use it at Stripe or Vercel in 2026. Those companies hire product designers who ship code. They expect you to own the surface end to end in the codebase not in design software. Do not use mock polishing when the market rewards speed and shipped product over perfect specs. The companies setting the pace proved that AI handles the mechanical polishing at equal or better quality. Your time is better spent on the decision making about which variant to ship and how it fits the larger product strategy. If polishing is still the bulk of your job run the audit in the death of the senior designer article immediately. Your role is in the collapse zone. The new senior spends less than ten percent of their time on anything resembling traditional polishing. The rest goes to shipping, deciding in ambiguity, and owning outcomes in code. Companies that still maintain a polishing step pay for work that a combination of Figma AI, v0, Galileo, Magician, Subframe, and Claude can complete in a fraction of the time. The economics are inescapable. Seniors who cling to this workstream will see their comp bands reset or their roles eliminated in the next reorg cycle. The shift happened faster than most predicted. Tools improved monthly throughout 2025. By early 2026 the old polishing role looked as dated as Flash design in 2012.

Mock polishing became the first senior task to fall because it followed discoverable rules that AI mastered quickly.