Two questions I get a lot are “how do I get some experience as a beginner?” and “how can I practice UX without a client?” Here are two suggestions, based on your experience. The “advanced” task is at the bottom.

(This is an edited version of an answer I posted on Quora.)

****

For Beginners:

This is something I recommend for any UX designer that wants to fill an empty portfolio and get into user research without too much risk or cost.

1. Download 10 professional apps in the app store. For best results, choose apps that are not the most common, but still well-known.

2. Create a short list of tasks that can be achieved in each app.

3. Find 5 people that are willing to give you some of their time. (You don’t have to do this all at once. Spread it out… people have shit to do.)

4. Get each of the 5 people to try each of the 10 apps, and do the tasks you decided on.

5. Do not help. Do not tell them what to tap. Even if they ask. Remember, they are testing the app, you are not testing them. Just tell each user the goal of the task/app and say “go”. Allow them to fail and figure it out.

6. Observe.

7. Make notes. (Record a video if you can!)

8. AFTER, ask them about why they did things the way they did them. Write down the choices they made, the paths they took, and what they believed (whether it makes any sense or not.)

9. Congratulations, you have now done 50 user interviews! If your first boss isn’t a UX designer, that’s probably more user interviews that they will have done in their career. But you’re not done yet.

10. Review your notes/videos and see if you can find some common mistakes, misunderstandings, flawed expectations, inefficiencies, etc. i.e. — look for problems that more than one user had in common. There is a good chance that part of the app can be improved.

11. Make wireframes for all the solutions you can think of. There can be 100 different ways to solve a simple problem. If you make 2 wireframes to solve/improve something in each app you will have 20 wireframes.

12. Make a portfolio where you treat each problem as a separate project. Explain the context, your research method, and your insights along with your solution (or solutions) for each set of wireframes. (More than one solution for a single problem is the reason we have A/B tests!)

13. Voila! You are now more experienced at UX than most people with a degree in HCI and you have a pretty solid portfolio to show for it. You’re the kind of intern or “first year” that I would invite to an interview without much thought, because obviously you care enough about UX and you understand what it takes to make designs user-friendly, and you’re smart enough to deliver.

And frankly: any boss who is experienced in UX will be able to teach you the rest, and any boss who can’t do UX will be so impressed you’ll be a serious candidate before you get to the interview.

****

For experienced UX designers:

This should intimidate you, because it’s difficult. That’s also why it’s good. Your task: redesign and specify Facebook.

Social networks are one of the most complicated types of websites on the internet, and Facebook is the biggest. If you screw up any of the “big chunks” of Facebook’s product, your mistake will be multiplied by millions or even a billion people, so it could cost Facebook an amount of money that will make your toes curl.

Here is what you need to include (the requirements) for your redesign. Note: I have never seen a “Facebook redesign concept” that achieves this list:

1. The Feed: is sorted using an algorithm called EdgeRank. It also includes many different types of content like videos, pictures, links, text posts, etc. Those posts can originally be created by you, or other users, or companies, and each piece of content can be visible to the public, friends, a group, a hand-picked list of people or just yourself.

2. Ad-based revenue: currently Facebook’s main money comes from ads. That means that engagement and page views are directly related to profit. Your re-design must consider ads a vital part of the design, while using all of the types of content named above. Bonus points: figure out a way Facebook can make billions of dollars without ads. ;)

3. Search: When you have a search on your site and there are a lot of different types of results (people, companies, groups, posts, etc.) it will be a project in itself to design the search and the search results page. In Facebook’s case, this is a crucial part of the navigation.

4. Feedback loops: One of the reasons it is so hard to design “social” is that you’re not just designing features, you’re designing human behavior, and it has to be viral. You don’t just have to create ways for people to follow, like, share, and discover content, you need to make sure that everything they follow, like, share, and discover is easily visible for everyone else. i.e. — if you share something and get zero likes, your engagement will be low.

5. Messaging: Anybody who has never designed a messaging app thinks it’s easy. Anybody who has designed a messaging app laughs at those people. Send, reply, forward, delete, archive, read/unread, groups, threads… however you make yours work, it’s not a detail.

6. Profile pages: These have been done over and over on the internet, and most are starting to look the same. Yours shouldn’t.

7. Improve something: Even if you achieve everything above, you will still only have matched Facebook. There is no point in re-designing just to be “new”. Your new design needs to be better in a way that can be measured. In Facebook’s case, it should make more money and get more engagement from users. Being “more beautiful” is not enough reason to redesign a site with a billion users.

8. If you’re a masochist: make it responsive, so everything in the list above works on the smallest mobile phone, the biggest desktop screen, and everything in between.

9. Wireframe everything, and write annotations for every feature. i.e. — number every button, piece of text, and link. Describe its function, where it goes when you click it, what happens when you hover over it, etc.

That list might sound rough, and it is, but as an experienced UX designer you need to try designing things that you have never seen before. Something with zero examples to copy. It forces you to consider the goals, the purpose, the users, the needs, and the limitations of what you’re trying to do. Even if you fail to solve it, the thinking you will do during the process is extremely valuable.

And for the record, I have designed several social networks from scratch. It’s hard, not impossible. ;)

Questions? Find me on Twitter: @JoelMarsh