Warm Gun: Design Process vs. the Real World at Instagram

by November 30, 2012

In his presentation at the Warm Gun Design Conference in San Francisco CA Mike Krieger discussed what can you take from the formal design process that will work within a start-up environment. Here's my notes from his talk on Design Process vs. the Real World Startup:

  • A true design process is still not a given in most organizations despite all the attention design has gotten these days.
  • In the design process you need to balance divergent and convergent activities. Getting the design right: where does this button go, how does interaction work, etc. Getting the right design: what are we doing and why.
  • Understand: get as much domain knowledge as possible. Do some primary research, don’t just use Google.
  • Observe: don’t just guess, watch. Do mini-ethnography out in the wild. At the end of this stage you want to know people better than they know themselves.
  • Distill insights: insights are things you are excited to tell others about. Aim for user + insight + design
  • Iteration: go for quantity, get your ideas out. Brainstorming alone often works better than a group.
  • Prototype & test: every prototype answers a question. Prototypes have semantic value -make sure you are learning.

In the Real World...

  • In the real world, you prototype, observe, ideate, and observe in real time. It’s the values of design school in a hectic, fast approach.
  • Draw on experience and previous understanding. Find where you know things and go from there. You don’t have time to start from no knowledge. Build on the years of background thinking you have already done.
  • Have a hypothesis on why you are different. Why are you doing this now? Have a point of view. Invert your assumptions: this helps you focus on how you are different.
  • Never code before sketching. Get the ideas out of your head and on to paper to decide if ideas are worth building. It’s ok if entire features get thrown out at the sketching stage.
  • What do you want to learn each week? Always begin with a question, work on that question for a week and see what you learn. Shipping every week help you build momentum.
  • Validate in social situations. The bar exam: if you can’t explain what you are doing to someone in a bar in 30 seconds, its too complex. Simplify.
  • Quick wizard of Oz prototyping. Fake out interactions manually to decide whether or not you should build them.
  • Know when its time to move on. Are there any remaining unanswered questions? If not, its time to move on to a new idea. If you know what question you are trying to answer, its easy to know when to move on.