Ask HN: How do developers learn design intuition?

4 points by mercat 251 days ago | 3 comments

As the title says: as a programmer and a wanna-be solopreneur, I often catch myself thinking way more about app architecture than about what actually matters for the user experience and satisfaction. Stuff like the API between frontend and backend, how the components are laid out, what data structures I’ll need, which questions the database queries answer. Which abstractions to introduce and how early in the development lifecycle, how to organize files so I do not confuse myself later.

On the flip side, when I build something, I am doing it for customer satisfaction. Any tool or product should solve someone’s problem. I can sketch a Problem Statement. And I can more or less orchestrate the code - how to capture the problem, solve it and pass the results along.

But showing it — that is the main problem for me. Yes, I can work in Figma. Yes, there are decent AI tools for generating designs now, but they mostly spit out the same thing. And a UI that looks average at best. It feels like as soon as I touch design, I start thinking it is not quite what I like and I often drift back to what I do like — app architecture, how everything connects and all that I already described. Wireframes are one thing, but when it gets to higher fidelity design, I end up leaning on pre-packaged UI tools and libraries like shadcn and similar stuff.

I would love to internalize at least a bit of the intuition that seasoned designers have, the folks who have done this for years and know how to lay things out.

I also get that this is like a designer asking a developer how to become a developer while knowing only 10% of what devs know and still shipping. But I will be a bit bold and say that visual design and the tools we have to create it still have fewer dimensions than software development.

I think it is also clear from my post that I have consumed tons of content and taken design courses, but nothing sticks (even after applying some to practice). If you are a developer who also designs, what advice would you give? Are there any foundational rules you can stick on the wall and lean on before you dive into a design from one angle or another? And how do you usually validate a design? Especially for views that are data heavy or where the user needs tools to get the job done, so they do not have to click too much to reach a goal.

ticulatedspline 251 days ago |

IMHO this is more like a designer asking another designer how to be a developer. In my experience most developers aren't great designers. I think that's why GUIs in open source software tend to be lackluster. those kinds of project attract developers, not designers and a GUI is often a second or even third class citizen to them.

While I've encountered devs that have a talent for both many had the "eye" before the code. While anything can be taught and learned if you have no proclivity towards it you're likely to never excel at it without pain. Imagine asking the same question if you were making music "hey devs who play instruments, how do I make good music". My advice would not be so much as "give up" but, "know your expertise", collaborate with someone who covers your weaknesses.

That said really good delivery is often hard won. different tasks have different concerns as well as different clientele. Get your designs in front of customers, there's software specifically for tracking UX metrics as well as general analytics. Don't trust that you can "know best", more than likely what you think is sensible someone else won't and given a large enough customer base people will flounder in your UI in ways that you could not possible fathom.

And most important: Copy whatever is already good. half of what they'll be doing is all the common tropes (gear means settings, hamburger means mo' menu), copy everything that works, users will find your product more naturally intuitive if they've already effectively learned it somewhere else.

mercat 251 days ago | |

> software specifically for tracking UX metrics

which UX metrics you’ve personally found the most valuable?

> Copy whatever is already good

it immediately reminded me of Steal Like an Artist. Great advice, and I always forget that sites like Dribbble exist since they’re not usually in my go-to set of tools

efortis 251 days ago |

Three tips:

Record the interactions (you, and with test subjects too). You’ll see more things.

Study competing UIs. Consciously reverse engineer their UIs. Don't know what you like but get the idea:

  - PhotoShop vs Gimp
  - Maya vs Blender
  - Excel vs Numbers
  - FCP vs Premier vs Resolve
  - Finder vs Explorer

Watch youtubers using them. For example, music composers have very interesting and different programs UI-wise. Or DJs, which need interfaces for live-performance.

Then give it a twist and use not directly competing programs that can be used for a similar goal. For example, for UI design:

  - Figma vs Illustrator vs Inkscape vs PowerPoint vs PhotoShop vs Word vs Excel vs plain text