How to Build a UX & Product Design Portfolio in 2026

A strong design portfolio is judged on how you think, not just on how things look. Reviewers spend most of their time trying to answer one question — can this person make good decisions under real constraints? — and a wall of polished screenshots doesn't answer it. Three or four honest case studies do. Lead each one with the problem and the call you made, not the final mockup.

That single shift — from showing outcomes to showing reasoning — is what separates a portfolio that gets a callback from one that gets a polite pass. Everything below is in service of it.

What design reviewers actually look for

When a design manager opens your portfolio, they are screening for a few things, roughly in this order:

• Judgment. What problem were you solving, what options did you weigh, and why did you choose the one you shipped? This is the whole game at mid and senior levels.
• Process, briefly. Enough of your method to trust your decisions — not a twelve-step ritual for every project. They want to see thinking, not a manual.
• Outcomes. What changed because of your work? Numbers if you have them; qualitative if you don't ("the support team stopped getting tickets about this flow").
• Range, then depth. Can you handle different surfaces and problems — and can you go deep on at least one? Show range across the set, depth within each piece.
• Craft. Yes, the visuals matter. But craft is the price of entry for a designer, not the differentiator. A beautiful portfolio with no reasoning reads as decoration.

The mistake most portfolios make is optimizing entirely for the last point and ignoring the first. Pretty is assumed. Thinking is what gets remembered.

Case studies, not screenshots

A screenshot shows the result. A case study shows the decisions that produced it — and decisions are what a reviewer is hiring. You don't need many; you need a few that are genuinely told.

A case study that works has a simple anatomy:

1. Context — what the product is, who it's for, and your role on the team. One or two sentences. Be honest about whether you led or contributed.
2. The problem — the actual problem, with its constraints. "Onboarding drop-off was high" is weak; "new users abandoned at the workspace-setup step because it asked for billing before they'd seen any value" is a problem someone can think about with you.
3. What you tried — the directions you explored and what you learned, including the ones you killed. Showing a discarded approach signals judgment more than a clean march to the answer.
4. The call you made — the decision and the tradeoff you accepted. Every real design decision costs something. Name it.
5. The outcome — what happened, measured if possible. If you left before results landed, say so and share what you observed.
6. What you'd change — one honest reflection. This is the most senior thing you can put in a portfolio, and almost nobody does it.

What to do about confidential or NDA work

This is the most common reason designers stall on a portfolio, and it has real answers. You can show your thinking without exposing anything you're not allowed to.

• Sanitize the visuals. Recreate key screens with placeholder data, swap the real brand for a neutral one, and blur or omit anything proprietary. The thinking survives; the secrets don't.
• Foreground process over pixels. You can almost always describe the problem, your method, and the tradeoffs even when you can't show the production UI. Whiteboard sketches, flows, and redacted wireframes carry the story.
• Keep results in ranges. "Reduced checkout abandonment by a low-double-digit percentage" respects an NDA while still demonstrating impact.
• Ask. Many companies will approve a specific, sanitized writeup if you ask the right person. The worst answer is no, and you're no worse off.
• Keep the deepest version off-site. Show range publicly, and when a recruiter wants the confidential detail, send it directly — a PDF, a Figma link, or a short deck. Your public portfolio stays NDA-safe; the sensitive depth goes only to people who ask for it.

Confidential work written up well often lands better than public work, because explaining impact you can't simply show forces you to communicate clearly — which is exactly the skill being evaluated.

How many projects, and which ones

Three to five. Fewer feels thin; more dilutes attention and buries your best work. Curate ruthlessly — a portfolio is an argument, not an archive.

Choose for range across the set and depth within it: one project that shows you can own something end to end, one that shows collaboration or a tricky constraint, one that shows a different surface or problem type than the others. If two projects make the same point, keep the stronger one. The piece you're proudest of goes first; reviewers often don't reach the third.

Where to host: your own site, or Behance and Dribbble?

Designers usually end up with both, and that's fine — but they do different jobs, and confusing them is a common mistake.

Concede the obvious: Dribbble and Behance are genuinely good for discovery, inspiration, and getting your shots in front of a community that's already looking. If freelance leads come to you through Dribbble, keep posting there. But a shot is a teaser, not a case study, and you don't control the layout, the URL, or whether it survives the platform's next pivot. Use the communities for reach; keep your definitive portfolio on a site you own and link out from there. Why that ownership matters in the long run is the subject of Own Your Profile.

The usual catch is that you trade one for the other — a community brings discovery, your own site brings control. Path is built to give you both: designed, portfolio-focused layouts and full case studies on a domain you own, with your profile indexed in a directory that hiring designers and recruiters search by skill and location. The home base and the surface area, in one place.

Make it load fast and get found

Design portfolios are image-heavy, which makes them slow by default — and a reviewer opening your link from a phone between meetings won't wait. Speed is part of craft here:

• Optimize every image. Compress, serve modern formats, and size them for how they're displayed. A good platform does this for you; if you build custom, it's on you.
• Write real alt text. It helps accessibility, helps search engines understand image-heavy pages, and is, frankly, a design-quality signal in itself.
• Use a clean URL and structured data so your site is the page that shows up for your name, with your role and location attached. The how-to is in How to Control What Shows Up When Someone Googles Your Name; the full build checklist is in How to Build a Portfolio Website in 2026.

Common questions

Do I need case studies, or are project shots enough?+

For most design roles above entry level, you need case studies. Shots prove you have taste; case studies prove you have judgment, and judgment is what's being hired. A practical split: lead with one or two strong visuals to earn attention, then deliver the reasoning underneath. Visuals open the door; the writeup is what closes.

I'm a junior designer with no client work. What goes in my portfolio?+

Self-initiated projects, redesigns of products you use, bootcamp or course work, and volunteer or community design all count — as long as you write them up with real problems and decisions. A thoughtful redesign of a checkout flow you found frustrating, with your reasoning shown, demonstrates exactly the thinking reviewers screen for. The best portfolio is the one that exists and shows how you think.

How long should a case study be?+

Long enough to show the decision, short enough to finish. A reviewer skims first, then reads if the skim is promising — so make it scannable with clear headers, lead with the problem and outcome, and keep the supporting detail tight. If someone can grasp the project from your headers and first lines alone, the length is right.

Should I include my visual design even if I'm a UX or product designer?+

Show the level of craft your role requires, and no more. A product or UX designer should demonstrate that they can produce clean, usable work, but doesn't need to perform brand-designer polish. Match the craft you show to the craft the job needs, and spend the saved effort on the reasoning.