I Like, I Wish, What If: A Better Way to Give Design Feedback

"Any feedback?" might be the worst question you can ask in a design review. It gives the reviewer a blank page, and blank pages get filled with whatever is easiest to say: nitpicks about the UI, vague praise, or a redesign of your work delivered as an order.

The best fix I know for this is the "I Like, I Wish, What If" framework, often credited to the Stanford d.school. It's incredibly effective because it forces the reviewer to balance their critique, and it removes the personal sting from negative feedback almost entirely. The conversation stops feeling like a performance review and starts feeling like collaborative problem-solving.

Here is how it breaks down, and how I use it.

"I Like...": the anchor

You start by highlighting what is actively working. This is not a compliment sandwich. The point is to explicitly validate good design decisions so the designer knows what not to change. It grounds the conversation in specific successes instead of vague praise.

Weak: "I like the colors."

Strong: "I like how the subtle background color groups the billing information together, it makes that section much easier to scan."

The weak version is a pleasantry. The strong version is information: it tells the designer that a specific decision landed, and why.

"I Wish...": the friction

This is where the critique comes in, but framed as a desire rather than a deficit. Saying "I wish" keeps the focus on the problem space and the user's experience, not on the designer's failure to account for something. It instantly lowers defenses.

Weak: "The typography on mobile is way too small and hard to read."

Strong: "I wish the mobile view accounted for users who might be viewing this outside in direct sunlight, the contrast feels a bit low."

Same underlying issue, completely different conversation. One is a verdict, the other is a problem worth solving together.

"What If...": the provocation

Finally, suggestions. But framed as hypotheses or experiments rather than direct orders. "What if" gives the designer a starting point for iteration without dictating the exact solution, which preserves their autonomy over the work.

Weak: "Move the submit button to the top right."

Strong: "What if we tried moving the primary actions closer to the user's thumb zone for the mobile layout?"

The weak version replaces the designer's judgment. The strong version invites it.

Why it works in practice

It kills the blank-page critique. When reviewers are just asked for "feedback," they default to whatever catches their eye first, usually surface-level UI details. The framework forces them to structure their thoughts: what's working, what's missing, what could we try.

It scales perfectly. You can use it in a live 1-on-1 design review, or you can have a team of five drop their Likes, Wishes, and What-Ifs onto sticky notes or a FigJam board for a rapid 10-minute asynchronous critique. The structure does the moderating for you.

It separates the problem from the solution. This is the part I find most valuable. "I wish" defines the problem. "What if" explores a solution. They are decoupled on purpose: if the designer doesn't like your "what if," they still have to solve your "I wish." The critique survives even when the suggestion doesn't.

Next time you're about to ask a room for feedback, try asking for likes, wishes, and what-ifs instead. The quality of what you get back changes immediately.