Thoughts on Critique

See Adam Connor and Aaron Irizarry, Discussing Design: Improving Communication & Collaboration through Critique, The book is at O'Reilly or at Amazon. The Discussing Design web site does not appear to be updated since Jan 2015.

For a good quick overview of design critique, have a look at the post by Cassie McDaniel, Design Criticism and the Creative Process. A List Apart is a web site about web design. The design process for web sites has overlap with Mechanical Engineering design in the areas of identifying customer requirements, the idea generation process and the importance of high-functioning, interdisciplinary teams.

On the inherent challenges in presenting your work

See Seth Godin' post on the paradox of the flawless record.

Basic goals

  • Be kind.
  • Be constructive.
  • Be a good listener.
  • Ask specific questions.
  • Look for the truth, even if the message is rude or painful.

Use "I" messages

  • Don't speak for anyone but you. Avoid phrases like "Everyone thinks …"
  • Use "I", not "you", especially in risky or potentially conflictual interactions
    • "I heard you say …" not "you said …"
    • "When I think about …" not "you should think about …"

It's not about what you "like".

Design critique is about the relationship between the work of your peers and their design objective. Your advice and opinion is being solicited. However, you are not being asked whether you like the work being done, or even whether you would use the same approach.

Of course, if your peer asks you directly, "Do you like this?" or "Does this appeal to you?", or "How would you solve this problem?", it is appropriate to respond directly. Just remember, if they don't ask for that kind of directive and personal opinion, it is better to ask them question and to focus on describing what you see in their design.

Before making comments, ask whether the team has identified specific problems or questions and invited your feedback on those issues?

"You asked me to comment on your design. Do you have any specific issues you want me to focus on?"

Then, when responding, keep these ideas in mind:

  1. First, consider the goals of the design. What is the team trying to achieve?

  2. Second, how does the work or ideas being discussed help the team achieve their goals.

Make observations, not commands. Framing the observation as a question can avoid redirecting the focus of the conversation back to your opinion and your preferences.

"I see that you have put a lot of effort into designing this mechanism. Is this the only way you have thought about moving the control head?"

"When I look at this idea, the first thing that comes to mind is my concern about safety of the sharp end of that beam. Do you have a plan for reducing that hazard? Or is this just a design concept?"

By asking questions you draw out the design team. You encourage them to look at their design with fresh eyes instead of forcing them to defend a decision that you judge as inferior. Taking this approach does not mean you withhold your negative reactions or opinions about the design.

In general, stay focused on the relationship between the team's work and their immediate and long term goals. Ask questions about how a feature or decision is helping the team achieve its goal. The reason for asking the question is that you see a mismatch between the goal and the methods of achieving that goal, or a potential weakness, or a problem that is interfering with attaining the goal.

The team may ask, "What would you do?" In that case, offering direct suggestions for how to act may be helpful.

Use a sandwich to deliver negative criticism

  1. State something positive
  2. Point out the weakness or error
  3. Reaffirm positive

Example:

  1. Your slide on the FEA results had a lot of good information
  2. I have a question about the boundary condition you used.
  3. Overall the results will help your design process

See also the post on the love sandwich from AIGA the Professional Association for Design.

To the receiver: Listen for the information, even it's poorly delivered.


Document updated 2018-02-12.