Most advice on improving design skills is vague. "Practice every day." "Study great design." "Get feedback." None of it is wrong, but none of it explains how.
How do you practice in a way that actually improves skill rather than just keeping you busy? How do you study great design in a way that transfers to your own work? How do you get feedback that changes how you design, not just how the specific design looks?
This guide answers those questions with the specificity that most resources don't provide.
Why Most Designers Stop Improving
The dominant mode of skill development for most designers is project work — you do client or employer projects, you complete them, you do more projects. This produces competence within your established patterns but rarely produces growth beyond them.
The problem: project work optimizes for completion, not learning. On a client project with a deadline, you use the approaches you already know because they're reliable. You don't experiment with unfamiliar techniques. You don't try risky solutions. You don't take the time to deeply examine why a particular approach works.
The result is a designer who gets better at executing their existing patterns efficiently, but who plateaus in terms of design quality, range, and insight.
Breaking this plateau requires deliberate practice — a specific type of practice that targets the edges of your competence rather than the center.
Deliberate Practice for UI Design
Deliberate practice is characterized by three things that distinguish it from regular practice:
1. It focuses specifically on what you can't do yet, not what you can already do well
2. It involves immediate, specific feedback on your performance
3. It's effortful — it doesn't feel like flow, it feels like working at the edge of your ability
For UI designers, deliberate practice looks like:
Recreating Great Designs
Choose a UI you admire and recreate it pixel-perfectly in Figma. Not copy — recreate, from scratch, making every decision yourself.
This is different from looking at inspiration. When you recreate, you confront every decision the original designer made. Why is the spacing 16px here instead of 12px? Why is this button outlined instead of filled? Why does the card have a border instead of a shadow?
The recreation forces you to understand those decisions. When you can't figure out why a decision was made, you learn the limits of your understanding. That gap is exactly where your skill needs to grow.
The discipline: Choose designs that are slightly beyond your current level — not so far that you can't figure out how to execute them, but far enough that they challenge your current approach.
Constraint Projects
Design the same interface under different constraints:
- Design a dashboard using only one color
- Design an onboarding flow with no illustrations, only typography and spacing
- Design a mobile app with no icons, only text
- Design a settings page where every piece of information is accessible in two taps or fewer
Constraints force you to find solutions you wouldn't have reached with full freedom. Some of those constrained solutions are better than what unconstrained design would have produced. All of them expand your design vocabulary.
Critique Practice
Find a design — from a product you use, from Dribbble, from Mobbin — and write a structured critique of it. Not "this looks good" or "I like the colors." A structured critique:
1. What is this design trying to accomplish? (the user goal and business goal)
2. What does it do well, and why? (specific decisions and their effects)
3. What could be improved, and how? (specific alternatives with reasoning)
4. What can you learn from this that applies to your current work?
Critique practice builds critical thinking about design decisions. Designers who can articulate why something works or doesn't work make better design decisions than designers who operate on instinct alone.
The Resources That Actually Improve Skills
Learn From Production Design, Not Portfolio Work
The single most impactful shift in how you learn from other designers: stop using Dribbble and start using Mobbin.
Dribbble shows you design that's optimized to impress other designers. Mobbin shows you design that's optimized to serve users — design that survived the constraints of real products, real development teams, and real user feedback.
The patterns in Mobbin are the patterns that actually work. Study them analytically: what decisions appear consistently across many products? What variations suggest product-specific constraints rather than universal best practice? What surprises you, and why?
Browse Learning Resources on Mantlr →
Refactoring UI — The Best Practical Design Book
Adam Wathan and Steve Schoger's Refactoring UI is the most immediately applicable design book available. Where most design books cover principles abstractly, Refactoring UI shows specific transformations — here's a design that looks amateur, here's the same design after specific changes, here's exactly what changed and why.
The specific improvements it covers — adding visual depth through layered backgrounds, improving hierarchy through font weight and size (not just color), creating space through intentional whitespace — are immediately applicable in the next design you work on.
Find Refactoring UI on Mantlr →
Untitled UI — Study the System, Not Just the Components
Using Untitled UI well is a design education in itself.
Open the Design Tokens page. Understand how the color scale is structured and why — why are there 9 steps? Why is the 600 step the "primary" step? Open the Typography page. Understand how the scale relates to itself — what's the ratio between sizes? Why are line heights different at different sizes?
Then look at how components are built from these tokens. Every padding value corresponds to a token. Every color corresponds to a scale step. The component is the visible output of a system of decisions.
Understanding the system rather than just copying the components is what turns a tool into an education.
The Shift That Changes Everything: Measure Your Work
Most designers improve slowly because they have no reliable feedback on whether their designs are getting better.
"This looks good" is not a measurement. "Users completed the onboarding flow at a 73% rate, up from 54%" is a measurement.
Find ways to connect your design decisions to measurable outcomes. In an employed role, request access to analytics and A/B test results. In freelance work, follow up with clients for data. In personal projects, set up Microsoft Clarity (free heatmaps and session recordings) and actually watch users interact with your designs.
The designers who improve fastest are the ones who know when their designs work and when they don't. This knowledge is only available from data.
The Mindset Shift That Separates Growing Designers From Stuck Ones
Most designers who plateau treat design as the production of artifacts — the deliverable is the file, the goal is completion.
Growing designers treat design as the production of outcomes — the deliverable is what happens when users interact with the file. The goal is understanding.
The difference in practice: a designer focused on artifacts says "the design is done" when the file is complete. A designer focused on outcomes says "the design is done" when we've validated that users can complete the task, the client's business goal is served, and we've learned something we can apply to the next project.
This shift changes how you approach every project. You ask different questions at the beginning (what outcome are we trying to achieve?), different questions during design (does this decision serve that outcome?), and different questions at the end (did we achieve the outcome, and what did we learn?).
It also changes how you improve. When you're learning from outcomes, you have clear feedback on what to improve. When you're only learning from artifacts, you only have aesthetic feedback.
The Weekly Practice That Builds Skill Consistently
Monday: Choose one design from Mobbin and write a 200-word critique. Be specific — identify three things that work well and three things that could be improved, with reasoning for each.
Wednesday: Spend 30 minutes recreating a component from a product you admire. Don't reference the original during creation — work from memory and your recreation will reveal what you don't understand.
Friday: Review work you produced three months ago. What would you do differently now? What decisions look naive? What looks surprisingly good? This retrospective reveals growth you can't see while you're inside it.
That's 90 minutes per week of deliberate practice, separate from your project work. Consistent over six months, this practice produces more growth than years of project work alone.