Think back to elementary school when you first learned about the scientific method. Here it is for those needing a refresher:

The scientific method
Image borrowed from sciencebuddies.org.

This is exactly what we do as user experience professionals. We build hypotheses, test hypotheses, analyze results, and draw conclusions. It’s how we prove whether our design solutions are achieving their goals. All of these steps are absolutely necessary in a successful UX process.

Let’s break it down, step by step:

1. Ask a question

This step could take many forms. It could be a question about how to improve an existing page (e.g. “How do we improve conversion rate on product page X?”). This question could also be about a totally new product (e.g. “How can we facilitate efficient point-of-sale transactions on a mobile device?”).

Think of this question as the goal of your project or the problem you are attempting to solve with your design solutions.

2. Do background research

Before you open Photoshop or even start wireframing, you need to go out and learn about the problem you are trying to solve. Learn as much as you can about the project’s subject matter. Look at competitors or similar products/sites. But even more important than any of the that, you have to go out and talk to stakeholders. This includes owners of the product or idea, current users (if applicable), prospective users, competing products’ users, and anyone else who may be involved with or affected by the end product. Learn about their needs and current pain points. Stakeholder groups will have varying perspectives/opinions, and this information will be invaluable as you construct a design solution.

3. Construct a hypothesis

This is your design solution, and it is based on the research you did in the previous step. It doesn’t have to be the perfect, final solution. In fact, it likely won’t be on your first pass, and that’s perfectly fine. You’ll have a chance to iterate later.

4. Test with an experiment

User testing! Whether you test it with five people around the office or 30 via a service like usertesting.com, you need feedback on your designs. This is where you verify whether or not your solution is really solving the problem you were tasked with in step one.

5. Is the procedure working?

Are your user tests really testing what you need them to? Isolating the pieces of your hypothesis that you want to test can be tricky. It may take a few tries to get it just right. Repeat steps 4 and 5 as needed.

6. Analyze and draw conclusions

Look at the results of your user testing and determine whether your solution is achieving what it’s meant to achieve. If it is, job well done! Go on to step 7. If they don’t, it’s back to the drawing board to create a new hypothesis (step 3).

7. Communicate your results

In other words, build it! Whether this means creating a detailed spec to send off to a development team or actually coding it yourself, you know you have a successful, user-tested solution ready to be let into the wild.


You absolutely need all of these steps to guarantee success when designing a user experience. Resist the temptation to skip the background research or testing steps as they are the only way to validate (or invalidate) the successfulness of your design. Without that, you’re just guessing.

I will leave you with this, from the great Neil deGrasse Tyson:

'Test ideas by experiment and observation. Build on those ideas that pass the test. Reject the ones that fail. Follow the evidence wherever it leads, and question everything. Accept these terms and the cosmos is yours. - Neil deGrasse Tyson'