We've launched Product Challenges

We've got a big update this week that I'm excited to share: we're launching a new challenge format called Product Challenges.

With our existing challenges (which we're now calling Design Challenges to distinguish the two formats), you implement a design as accurately as possible. You get a Figma file (or JPG mockups), and the goal is to match it. Product Challenges takes a different approach. There's no Figma file. You get a product spec, and you make the design decisions yourself. You work with AI, build the full product, and ship something that real people can use.

Design Challenges aren't going anywhere, and they already cover a huge range, from small components up to full-stack apps. Product Challenges are a new format alongside them, designed to help you build a different set of skills.

---

What you get (and what you don't)

Instead of a Figma file, you get a product spec. Here's what that includes:

• A product definition with who it's for, what it does, and why it matters
• 15-20+ features organized into Core and Stretch tiers, with clear acceptance criteria
• 2-3 open-ended features where you get a problem statement instead of a solution. You research, decide, and explain your approach
• A brand kit with colors, typography, spacing, mood, and links to real products for inspiration. Think of it as a strong starting point, not a rigid prescription. You can also skip it entirely and create your own brand direction
• AI context files that give tools like Claude Code, Cursor, and Copilot full project awareness from the first prompt
• Sample data with real-world edge cases, the kind of messy, inconsistent data you'd actually see in production
• A design concept image showing one possible direction for the main app screen, plus links to similar real products for inspiration. Use it as a starting point, take it in a completely different direction, or discard it entirely
• Differentiators: 4-6 enhancements across different skill areas. Pick 1-2 to make the project yours

What you ship feels like a real product. You build a landing page that explains what your app does, and a guest experience that lets anyone explore the full app without creating an account. These are products you could actually use day to day, share with friends and family, or even put in front of strangers.

We recommend building them as full-stack applications to get the most out of the format. But every Product Challenge also includes a frontend-only path with sample JSON data, so if you're focused on frontend skills, you can still take on any of these challenges and practice the product thinking, design decisions, and AI collaboration without needing to set up a backend.

---

Why this format, and why now

The role of a software engineer is changing fast. Full-stack roles are growing while frontend-only postings are shrinking: between January 2024 and January 2025, global job postings for frontend developers fell 25% while full-stack postings grew 9%. Entry-level positions across the board have dropped roughly 40% from their 2022 peak, and the roles that remain increasingly expect broader skills. Figma's own research found that 55% of product builders are taking on tasks outside their usual scope, and 72% say AI tools are the main driver.

Companies increasingly want product-minded engineers: developers who don't just take a spec and implement it, but think critically about what's being built. As Gergely Orosz noted recently, being product-minded is becoming a baseline at startups because with AI handling more of the implementation, engineers need to be the ones directing what gets built and how. The spec tells you what features to build. It doesn't tell you how the interface should work, what the best experience is for the user, or how to handle the edge cases that only surface when you're deep in the build. That's where your judgment comes in.

The day-to-day work is shifting. Engineers are spending less time writing code and more time orchestrating AI agents that write code for them. At Atlassian, some teams have engineers writing zero lines of code directly, with everything handled through agent orchestration. The SF Standard recently reported that engineers are being asked to think more like product managers, and product managers are expected to be more technical, with the lines between the two blurring. The industry is moving toward spec-driven development, where engineers work from detailed specifications and collaborate with AI agents to generate code, rather than implementing detailed design mockups. The code itself is becoming the easier part.

The best way to develop these skills isn't by reading about them. It's building things: going back and forth with AI across multiple sessions, figuring out the workflows, discovering where AI is strong and where it falls short, reviewing what it outputs, and learning to steer it toward good results.

Product Challenges are designed around exactly this. You work from a spec. You operate within a brand kit (like you would a design system). You collaborate with AI to move fast. And you bring the judgment that AI can't provide on its own: the product decisions, the design sense, and the attention to detail that make software genuinely good.

---

The skills every challenge develops

Product thinking

Some features in each challenge are fully specified. Others give you a direction and leave the rest to you: "Users want insights from their data. Design what to show and how to present it." That gap is intentional. In real jobs, specs rarely spell out every detail. Being able to take a vague requirement, make a good call, and explain why is something companies really value.

Design taste and craft

Without a Figma file, your design sense is on full display. The brand kit gives you a starting point, but the layout, component styling, animations, how you handle empty states, loading, and overflow: that's all on you. The gap between "it works" and "it works and looks great" is exactly what employers notice.

AI collaboration

Product Challenges are designed with AI collaboration in mind. Context files give AI tools the full picture of the project, the spec, the brand kit, and how to collaborate. The idea is that AI handles the speed of implementation while you direct the product decisions, design quality, and edge cases. Working well with AI is a real skill now. With roughly 85% of developers using AI coding tools regularly and most juggling two to four tools at once, it's become something you need to practice, not something you pick up on day one of a job.

Shipping real products

Our Design Challenges can be built as full-stack apps, and many of them are. What Product Challenges add is the product layer on top: a landing page that communicates what the app does and why someone would want to use it, and a "Try as Guest" experience so anyone can explore the full app without signing up. It's a chance to practice and showcase a different set of skills: selling the product, thinking about first impressions, and making something that feels complete from the moment someone lands on it.

---

Three challenges at launch

We're starting with three, each with a different flavor.

Frontpage: RSS feed reader (free)

Design concept for the RSS feed reader product challenge

Build a content aggregator that pulls RSS and Atom feeds into a clean reading dashboard. Your personal front page for tech content.

You design the onboarding flow, the digest view, and the layout system. The guest experience comes loaded with 19 feeds across frontend, design, backend, general tech, and AI, all real content from real sources.

The engineering challenge here is fun: RSS has been around since 1999, and the real world shows it. You'll deal with encoding inconsistencies, broken XML, wildly different date formats, and feeds that give you full articles alongside feeds that give you one-sentence summaries. It's the kind of messy, practical work that tutorials never cover.

Frontpage is free, so anyone can start today.

Bookshelf: personal reading list (premium)

Design concept for the personal reading list product challenge

Build a reading tracker that lets you search for books via an API, organize them into shelves, track progress, set goals, and explore year-in-review statistics.

You design the year-in-review screen, the book-discovery flow, and the reading-progress UX. The guest experience has 45 real books across 5 genres with real covers, ratings, notes, and progress data.

The year-in-review is one of the challenge's design-it-yourself features: you get the problem ("users want to reflect on their reading year") and you decide what to show and how to present it. Books per month, genre breakdown, total pages, reading pace, all designed by you as a distinct visual moment. It's the most shareable part of the whole project. And the API integration is real-world messy (missing covers, multiple ISBN formats, duplicate editions, incomplete metadata), which is what working with external APIs actually looks like.

Skills Learning Tracker: practice journal (premium)

Design concept for the skills learning tracker product challenge

Build a practice tracker where you log sessions, manage skills, track streaks, and visualize your consistency over time.

You design the dashboard and the session logging experience. Guest mode has 6 skills (Spanish, guitar, cooking, TypeScript, AI engineering, and UI design) with 47 logged sessions, active streaks, and months of heatmap data.

There's a lot of room for creative design decisions here: how you visualize streaks, how you present consistency data, how the dashboard comes together. It's also the most focused in scope since there's no external API to wrangle, which makes it a great place to really dig into your AI collaboration workflow and iterate on the design. If this is your first Product Challenge, start here.

---

A practice ground and a proving ground

Our mission at Frontend Mentor has always been to help you build the practical skills you need to succeed as a software developer. Whether that's implementing designs accurately with clean, accessible, responsive code, or taking a specification and building a fully featured product from it. We want you to be a confident developer from day one, and to keep growing into an excellent mid-level developer and beyond.

As the industry shifts, we want our formats to shift with it. Design Challenges provide the best practice ground for design-led development, and that's not going anywhere. Product Challenges are our response to the move toward spec-led development, AI collaboration, and product-minded engineering. As the way software is built continues to evolve, we'll keep adapting what we offer so that Frontend Mentor remains the most realistic and relevant proving ground for software developers.

Whether you're working toward your first role, you're a junior developer looking to level up, or you're more experienced and want to sharpen your product and AI collaboration skills, Product Challenges are here for you. They're available at intermediate, advanced, and guru (coming soon!) levels.

If you're not yet comfortable with Design Challenges at the intermediate level or above, we'd recommend sticking with those for now and continuing to build your fundamentals. Working through different designs exposes you to different design styles, UI patterns, and problem-solving approaches, and that foundation will serve you well. Once you feel solid, Product Challenges are here to help you build experience and skills with the spec-led workflow many professional teams are adopting.

These three challenges are the start. Product Challenges will now be part of our regular challenge launch schedule alongside Design Challenges. You can still expect new Design Challenges with Figma files and detailed designs, but you'll also see new Product Challenges landing at a regular cadence.

---

Start building

Product Challenges are live. Pick the one that interests you and go.

• Frontpage — RSS feed reader. Free. 12 core features, 5 stretch. Design the onboarding, the digest view, the layout system. Parse real feeds. Ship something you'd actually use.
• Bookshelf — Reading tracker. Premium. 12 core features, 7 stretch. Hook up a book API, design the year-in-review, track your reading. Ship a library you'd come back to.
• Skill Learning Tracker — Practice journal. Premium. 9 core features, 5 stretch. Design the dashboard, visualize your consistency, track your streaks.

Give them a go and let us know what you think. We're excited to see what you build!