As designers we all shift between pushing pixels/coding to thinking about big picture things — how do you manage this shift?

By maintaining and reviewing an active list of personal and professional goals. Meditation. Exercise. Stepping away. Is that what you mean?

Ack. Sorry. I mean big picture ideas for the product. For example I've been spending time building parts of the UI and coding, and the next project is thinking about how to make it easier for people to get their ideas from their head onto Branch. I keep getting sucked into UI details like making compose box smaller, but that's too low level.

I usually put my thinking hat on by stepping away from the computer and sitting on our sofa with a notebook. I
don't even sketch wireframes right away, I make lists of ideas. Unless I keep sketches suuuuper scribbly, I get sucked into the "UI design" hat.

Defining a clear problem statement (and making sure it's the right problem) also helps, otherwise it's difficult to judge which solution is the better one.

Yeah, I'm still with my original statement on this, and just seconding Cemre's. I know it's a trite saying to state that design is problem solving, but it's true. You say the solution is making it easier for people to get ideas from their head into Branch, but what's the problem? I have no problem typing this. Define the problem, deconstruct it.

Also, awesome discussion, thanks for starting it. Right now my biggest design problem at work is the process in which we nurture and develop ideas, so this is topical.

Also, whenever our CEO Josh pitches me an idea about something we should build, instead of lazily accepting that it's a good idea and starting to design the UI, I try to play devil's advocate and come up with reasons why we shouldn't do it. Doing this forces us to have a higher level conversation and exposes what's the root of the idea. Once we identify that root (usually a problem), I feel more confident thinking about implementation.

Shut down the computer and go out with a pen and notebook. Then look back at core objectives of the product and problems. I make a lot of lists. Things we need to do long term. Ideas I would love personally.

Does the latest pixel perfect mock still solve the core problem? Is it introducing new ones? Often I like to just go out and do something totally different and not think about the design at all. A day or two later I look back at the designs and tend to see lot of loopholes or confusing states.

Another thing I do a LOT is think of the product from two perspectives, a tech noob and a totally new user who does not understand the interface language. Are things still obvious for them? Do they need any nudging and helpful guidance?

One of the biggest things I have seen happening is we get so close to the feature and its nuances that it is easy to mistake complexity and obviousness. Often I learn a LOT by showing stuff to someone who knows nothing about the stuff I am working on. Sometimes the simple question, "WHY are you guys doing this?" leaves me speechless since the original goals kind of get lost during the process.

Also if the company culture is "build on top of other ideas and always say AND over NO" it is easy to get trapped into a spiral of "what if we do this". The original objectives and long term goals list ALWAYS helps me bring everyone on the same page, and many times that includes me.

Also struggling with this by wanting to jump onto the details way too early. Here what I "try" to combat that:

+1 on stepping away and sketching more
+ Keeping a "wireframe" style for as long as possible. This means, no colors, no shadows, just flat shades of grey until there is a working prototype and you can test user interaction. Then at the end you can start painting on top. The only problem, that works fine for a 1.0, but after it's too late. So what I'm gonna try next is to keep the "wireframe" styles and the "paint" styles separate so you can easier add/remove the skin.

I'm terrible at keeping the layers separate — brand/visual/UI/UX/product blend together. But I have been surprised to find two elements to my process that are crucial in a way I never would have expected.

Ambiance — I find that product design in particular is more of an ambient problem solving process. It doesn't seem to come as much by careful and logical consideration. Of anything I do, product design feels the closest to a "gut" thing. My co-founder works more from logic and when we work on product it's a nice complimentary process. But overall, I don't shut this process off ever. Even the most minute details can have broad psychological impact for users so I need to be aware of how the pieces work together.

Words...

Words — language plays a huge role in my process. This is common for the higher level problems. We write out stories, flows, profiles, and theses to help make sense of a problem. But I often reason through lover level problems like interactions by writing out a description of them in a very conversational tone, proposing ideas and asking questions.

I've gotten some surprised reactions to this. I tend to sketch and draw less and instead use language before to design.

The second half of using language is discussing the work. So much of our forward momentum comes from conversation over the work. It's a way to feel through the dark in a process of discovery and find unexpected solutions. I wrote about this process here: blog.keenancummings.com