10 Step UX

Every designer does it differently, but it’s always interesting to read how one workflow varies from another. Figured I’d share mine. Enjoy!

1. Research

Research is your holy grail. Include things like competitor analysis and moodboarding here. You’re aiming for less but better; what can you learn from your competitors? Where can you reduce friction? The most important question to be asking yourself during the research phase is: do you really need to reinvent the wheel? I spend at least 30% of my time doing this. It shapes so much future ideation of a feature that you can’t afford to skip it.

2. Data Analysis

Not always a necessary step, but usually a very useful one if utilised correctly. Data-driven design is all the rage nowadays, and there’s good reason. Anything from Mixpanel to Intercom can give you further insight into some of the research you might have done.

3. User Personas

Consider this a prerequisite. You should always have your users in mind — but it doesn’t hurt to have a look over some personas now and then.

4. IA and Considerations

By now you’ve drawn a conclusion of what might be where, but think about how data is displayed and when? Each content level can be displayed in a different way — consider each method of articulation a bullet spent from your chamber — you can’t reuse the same design elements. Structure is king here.

5. Lo-fi

My favourite time of day — time to hit the whiteboard. Work through low fidelity screens and refine, refine, refine. Ensure you have a sounding board. Managers, product people, snakes and rubber ducks all suffice. Ever heard of rubber duck debugging? Tell yourself user stories, sometimes explaining something is all you need to fix it.

6. Hi-fi

Sketch, Photoshop, slab of granite — whatever works. Just get the flows looking like they would on the end platform. Skip wireframing with grey boxes and shit, it’s a waste of time.

7. User feedback

Given your research and analysis got you to some good conclusions this bit should be nothing but touching base with users to ensure you’re going along the right lines. No testing just yet. Formulate questions that will get opinions, this can even be internal.

8. Prototype

You can prototype in anything from Xcode to Framer.js. But if, like many people, you’d prefer to punch yourself in the face repeatedly than write some Javascript you can use tools like Flinto or Principle.

9. Microcopy and microinteractions

Got a bitchin’ bounce animation? Put it together after the prototype is completed. Users like to feel acknowledged, and these things add to that.

10. Test

Your users tell you what you’ve built is worthless, spend two days in nihilistic dread.