A list of one hundred project ideas sounds like the solution to the problem.

Until you have to choose one.

Suddenly, a medication management app feels responsible and socially meaningful. A platform for local artists seems more creative. A sports facility booking system sounds realistic. An energy consumption dashboard could demonstrate more advanced skills. An app for an animal shelter has heart. An administrative system has a table.

Half an hour later, you still do not know what to design, but you have already become emotionally invested in six fictional startups.

Choosing a portfolio prompt should not be about finding the idea that sounds most interesting. It should begin with a much less glamorous question:

What should this project prove about the way I think and design?

If you are still looking for a topic, start with my collection of 100 UX/UI Design Prompts. This article will help you choose a project that can genuinely become a valuable case study—instead of another set of polished screens nobody needed.

A Good Prompt Is Not Yet a Good Project

A prompt gives you a starting point:

Design a platform that helps small organisations manage volunteers.

That is enough to open Figma.

It is not enough to create a strong portfolio project.

The prompt does not yet define:

• who exactly will use the product,
• which problem matters most,
• why existing solutions are insufficient,
• which constraints should shape the decisions,
• what should be designed first,
• what should deliberately remain outside the scope,
• which skills the project is supposed to demonstrate.

The same idea can lead to completely different outcomes.

One person may design a mobile app with a calendar, a volunteer profile, and colourful status labels. Another may focus on assigning people to tasks, availability conflicts, required qualifications, and communication between the coordinator and the team.

The first person shows a screen.

The second shows a product problem.

Do not choose a prompt only because of its subject. Choose it because of the decisions it will allow you to demonstrate.

1. Start With the Skills You Want to Prove

A common mistake looks like this:

I want a fitness app in my portfolio because it will look good.

That is not yet a portfolio strategy. It is an aesthetic preference wearing sportswear.

Instead of asking what product you want to design, ask:

What skill should a recruiter or client recognise?

It might be:

• simplifying a complex process,
• designing forms and validation,
• building information architecture,
• working with large amounts of data,
• onboarding new users,
• designing a permissions system,
• handling errors and exceptional states,
• accessibility,
• designing for multiple roles,
• making product decisions,
• creating a coherent interface.

Then choose a prompt that naturally requires that skill.

Example

If you want to demonstrate the design of complex processes, a simple landing page for a coffee shop will probably not help.

A better project would include:

• different statuses,
• conditional decisions,
• user roles,
• interrupted processes,
• data requiring verification,
• actions dependent on previous steps.

On the other hand, if you want to show visual sensitivity, hierarchy, typography, and brand building, an extensive administrative platform may only increase the scope unnecessarily.

Not every project has to prove everything. Trying to demonstrate every possible skill in one case study usually ends with a product that has a marketplace, a community, chat, an admin panel, artificial intelligence, and probably its own currency.

2. Match the Project Level to Your Experience

The project should be neither too simple nor absurdly broad.

For a Beginner

A good prompt should allow you to focus on:

• one main problem,
• one primary user group,
• one key flow,
• a limited number of screens,
• a clear goal.

Example:

Design a system for booking study rooms in a local library.

You can focus on:

• room availability,
• selecting a date and time,
• booking rules,
• cancellation,
• confirmation.

That is enough to demonstrate process design, states, validation, and basic information architecture.

For a More Experienced Designer

The project may include:

• multiple roles,
• complex dependencies,
• system-level consequences of decisions,
• data analysis,
• configuration,
• scalability,
• integrations,
• failure scenarios.

Example:

Design a platform that allows a network of libraries to manage spaces, bookings, and access policies.

The subject may sound similar, but the problem is much broader. It introduces:

• branch administrators,
• different types of spaces,
• local policies,
• booking conflicts,
• permissions,
• reporting,
• exceptions.

Do not choose a complex product simply to appear more senior. Seniority is not about adding six user roles and a dashboard. It is about managing complexity deliberately.

3. Choose a Problem You Can Understand Credibly

A fictional project does not have to be superficial. It does, however, need to be based on accessible context.

Before choosing a prompt, check:

• whether you can speak with potential users,
• whether you know someone working in the industry,
• whether reports, forums, or other materials are available,
• whether you can analyse existing solutions,
• whether you understand the basic constraints of the domain,
• whether you can separate facts from your own assumptions.

A project for students will be easier to research if you are a student or have access to people who study.

A system for surgeons may sound fascinating, but if your only sources are a medical drama and three Reddit posts, you have a problem.

This does not mean you should design only for people exactly like you. It means the scope of the project should reflect the quality of the information available.

Particular care is required in areas such as:

• healthcare,
• finance,
• safety,
• law,
• disability,
• violence,
• mental health crises,
• migration,
• public services.

These are not forbidden topics. They are simply areas where a polished interface cannot compensate for a shallow understanding of the consequences.

4. Check Whether the Project Contains Real Decisions

A good prompt should lead to questions that do not have one obvious answer.

For example:

• Should the user create an account before starting the process?
• Which information is mandatory?
• When should the system block further progress?
• Should data be shown immediately or progressively?
• What can the user do when the process fails?
• Who can edit, approve, or delete information?
• How should speed be balanced with safety?
• What should appear in the default view?
• Which features are genuinely necessary in the first version?

If the project is limited to choosing fonts, colours, and card layouts, it may be a useful UI exercise. It will not, however, become a strong UX case study.

Weaker Prompt

Design a modern weather app.

Stronger Version

Design a weather app for people who work outdoors and need to assess risks related to heat, thunderstorms, strong winds, and air quality quickly.

The second prompt immediately introduces decisions:

• which data should have the highest priority,
• how risk should be communicated,
• how to avoid information overload,
• when alerts should be sent,
• how to design for someone using a phone in difficult conditions.

The subject is still weather. The difference is that it now has a specific context of use.

5. Narrow the User Group

“An app for everyone” usually means an app designed for nobody in particular.

You do not need an elaborate persona with a favourite coffee brand and a quote about technology. You need to understand who is in a particular situation and what they are trying to accomplish.

Instead of:

a platform for jobseekers

choose:

a tool for people returning to the job market after a long career break

or:

a system that helps candidates manage several recruitment processes at once

or:

a platform for junior designers preparing their first portfolios and job applications.

Each of these groups has different needs, fears, and behaviours.

Narrowing the audience does not limit the project. It gives it direction.

6. Choose a Realistic Scope

A prompt may sound harmless:

Design a platform that supports local communities.

In practice, it may involve:

• events,
• groups,
• listings,
• messages,
• a marketplace,
• fundraising,
• organisation profiles,
• moderation,
• safety,
• maps,
• recommendations.

That is not a prompt. That is several years of roadmap and the first disagreement between the Product Manager and the finance department.

Instead of designing the entire ecosystem, choose one part:

Design a process that allows residents to report local problems and track their resolution.

Now the scope may include:

• creating a report,
• location,
• photos,
• category,
• status,
• updates,
• providing additional information.

The project becomes possible to complete and describe.

A Simple Scope Test

Try describing the project in one sentence:

I help [a specific group] complete [a specific task] in [a specific context].

If the sentence requires three commas, two sets of brackets, and the phrase “as well as”, the scope probably needs further reduction.

7. Avoid a Project That Is Only a Copy of a Popular Product

You can design a streaming app, a banking app, or a food delivery platform. The popularity of the subject is not the problem.

The problem appears when the project looks like this:

I copied Spotify, changed the colour, and called it a UX case study.

A conceptual project should have its own problem, context, or constraint.

Instead of:

redesign a music app

try:

design a way to discover music for people who feel overwhelmed by recommendation systems and want to explore new genres more intentionally.

Instead of:

a food ordering app

try:

design a meal ordering process for a group of people with different dietary requirements placing one shared order.

A familiar product model can be an advantage. It allows you to compare solutions and understand basic mechanisms more quickly. You still need to find your own problem rather than simply recreate someone else’s interface.

8. Do Not Design Around a Screen You Want to Show

Sometimes the choice of prompt begins with:

I want a dashboard in my portfolio.

A dashboard is not a user problem. It is a format for presenting information.

First, you need to understand:

• who needs the information,
• which decisions they make based on it,
• how often they use it,
• which signals require action,
• what represents a normal state,
• what indicates a problem,
• which actions can be performed.

Only then may it become clear that a dashboard is the right solution.

The same applies to:

• maps,
• chatbots,
• AI recommendations,
• calendars,
• communities,
• reward systems,
• configurators.

Do not choose a project because it will allow you to design a fashionable component. Choose a problem that can justify using one.

9. Evaluate Whether You Can Show the Process Without Inventing It

A fictional project is perfectly acceptable.

Fictional business success is not.

If you do not have access to real data, do not write that:

• conversion increased by 34%,
• completion time was reduced by half,
• the solution increased retention,
• users became more satisfied,
• the company reduced costs.

You can show instead:

• which risk the solution was intended to reduce,
• which assumptions required validation,
• which decisions you made,
• which alternatives you rejected,
• which tests you conducted,
• what changed after feedback,
• what you still do not know,
• what the next validation step would be.

A good case study does not have to end with a triumphant infographic and a green arrow. It can conclude honestly:

The solution simplified the main flow in the prototype, but it would require further validation with users and the technical team.

It sounds less spectacular.

It also sounds more credible.

10. Choose a Subject You Are Willing to Develop for Several Weeks

A portfolio project requires more than initial enthusiasm.

Before selecting a prompt, ask yourself:

• will the subject still interest you after completing a competitor analysis?
• do you want to read about the domain?
• can you discuss the problem without returning only to the appearance of the screens?
• are you curious about the potential users?
• does the problem offer room for iteration?

Do not choose a project only because it looks good on Behance.

If, after two days of research, the only part that still excites you is the gradient on the login screen, it may be time to return to the prompt list.

A Simple Method for Evaluating a Prompt

Score each idea from 1 to 5.

1. Alignment With Your Portfolio Goal

Will the project demonstrate the skills required in the roles you are applying for?

2. Access to Knowledge

Can you obtain credible information about the users, the problem, and the domain?

3. Decision-Making Potential

Does the project require meaningful trade-offs, prioritisation, and work with constraints?

4. Realistic Scope

Can you complete it without designing an entire digital empire?

5. Case Study Potential

Will you be able to show the process, decisions, iterations, and conclusions?

6. Personal Interest

Is the subject interesting enough that you will not abandon it after designing the hero section?

The maximum score is 30.

This is not a scientific model. Do not send it to a laboratory. Its purpose is simply to make you compare ideas using something more useful than “this screen could look nice”.

Example: Three Different Prompts

Imagine you are considering three projects.

A. A Travel Planning App

It could demonstrate:

• search,
• information organisation,
• planning,
• maps,
• collaboration.

Risk: enormous scope and intense competition among portfolio projects.

B. A Medication Management Tool for Older Adults

It could demonstrate:

• accessibility,
• reminders,
• safety,
• readability,
• critical errors.

Risk: a difficult domain requiring careful handling of medical considerations.

C. An Equipment Booking System for University Students

It could demonstrate:

• resource availability,
• statuses,
• scheduling conflicts,
• permissions,
• collection and return,
• penalties or restrictions.

Risk: visually less glamorous, but with excellent process-design potential.

The best choice depends on the goal.

If you want to demonstrate accessibility, project B may be the strongest.

If you are applying for B2B or enterprise product roles, project C may say more about you than another beautiful travel app.

A good prompt does not need to sound exciting at a party. It needs to give you material for demonstrating mature design thinking.

What to Do After Choosing a Prompt

Do not open Figma immediately.

First, write down:

1. Who has the problem?
2. What exactly is the problem?
3. In what context does it occur?
4. How does the user handle it now?
5. What information are you still missing?
6. Which part of the problem matters most?
7. What is deliberately outside the scope?
8. Which skills do you want to demonstrate?

Then follow a simple sequence:

prompt → problem → user → context → constraints → decisions → solution → case study

I will explore this process in a separate article: From UX Design Prompt to Portfolio Case Study.

Where Does ChatGPT Fit Into All of This?

AI can help you:

• organise notes,
• identify assumptions,
• prepare research questions,
• find potential edge cases,
• compare directions,
• create a documentation structure,
• practise presenting the project.

It should not:

• invent users,
• generate fictional research findings,
• make decisions for you,
• write a case study about a process that never happened.

In 15 UX Prompts for ChatGPT That Support Real Design Work, you will find practical prompts supporting research, analysis, documentation, edge cases, and project storytelling.

The difference is simple:

A project prompt helps you choose what to design.
An AI prompt helps you analyse the work that begins afterwards more effectively.

The Best Project Is Not Always the Most Impressive One

Choosing a prompt is not a competition for the most innovative startup.

You do not need an idea that “nobody has ever done before”. In most cases, there is a reason nobody has built a platform combining healthcare, blockchain, artificial intelligence, and neighbourhood plant exchanges.

You need a problem that allows you to demonstrate:

• how you gather information,
• how you reduce scope,
• how you make decisions,
• how you work with uncertainty,
• how you consider consequences,
• how you change the solution when new information appears,
• how you explain your reasoning.

The screens are the result of that process.

Not a substitute for it.

Start With an Idea, but Do Not Stop at the Idea

You can browse 100 UX/UI Design Prompts and choose a subject that matches your level, interests, and portfolio goal.

Then ask yourself three questions:

Can I understand this problem credibly?
Will the project require meaningful decisions?
Will those decisions demonstrate the skills I want to be hired for?

If the answer to all three is yes, you have probably found a good starting point.

Not a perfect one.

A sufficiently good one to begin real design work.

---

Do You Want to Use AI While Developing the Project?

The Designer’s AI Playbook includes a library of 50 UX prompts, the Lead → Expand → Refine → Elevate framework, and practical workflows for research, documentation, decision analysis, collaboration, and portfolio creation.

The book does not help AI design for you. It helps you use AI to organise your thinking, explore alternatives, and prepare materials more efficiently—without surrendering your professional judgement.

Explore The Designer’s AI Playbook →