This is often overlooked, but very important. What are the elements on the page? What elements are most important?

If we don’t answer questions about the hierarchy early, it will be hard to defend decisions later. Everybody has different interests in a design, and it’s important to be able to point at something and say “this can’t be the most important thing on the page, and here is why.”

At a bigger company, this might be it’s own position. At Genius, everyone is thinking about information architecture. However, as the ones actually producing the designs, designers have to be the backstop!

Design discussion amongst the community can continue for days. Get started now, and consolidate the best feedback into a checklist tomorrow.

We don’t need beta testers at Genius, because our community is awesome at it. We use the Genius updates forum to keep people in the loop and get feedback.

A little bit of pseudo code here, explicitly:

At least one designer needs to review the designs. Designers at Genius know more about this stuff than anybody else on the team, so get their feedback!

Somebody with high-level perspective. Most projects involve more than design, so designers need to get feedback from somebody who knows about that stuff. Usually, this is the project backstop or @lemon.

I have friends who design hundreds of mockups for a project, nailing down every detail for every step in a flow. Everybody’s process is different, so mostly complete might mean anything. Let me clarify.

Mostly complete is just enough stuff to know what to do. Since designers at Genius implement the designs, communication overhead is low. We use this to get projects out faster.

This is very difficult for designers (myself included), but it’s also super important! By getting into review early, we can keep the project moving in the right direction and reduce unnecessary work.

This depends on the project, but preferably choose a safe place. At the 20% stage designs aren’t polished and sometimes don’t make sense. The people who should be seeing 20% mocks are people who understand the project and can fill in the blanks.

Sometimes the dream doesn’t happen quickly, or ever. If we rip out a feature and deploy, the replacement might ship 3 months later, leaving the community with a half-finished product in the meantime.

It’s better to wait until the replacement is ready, then swap.