A review rubric is the list your company checks when it decides how good you are at your job.
Most design rubrics are a year out of date. In this year's AI in Design survey of more than 900 designers, 73% said expectations on output, quality and speed had gone up. Only 13% of design leaders had updated their review metrics or their hiring process to match.
That gap is not a lag that closes on its own. It is a rubric that has quietly started rewarding the wrong work.
AI did not raise the bar. It deleted the thing the bar was measuring.
Five ways the old rubric now works against you
For each one: what your review still asks for, why that broke this year, and what to put in its place.
1. Volume stopped being proof
What the rubric says: more screens, more flows, more iterations.
Why it breaks: Weekly AI use for design work went from 54% to 91% in a year, and three quarters of designers now use it daily. Iterations are close to free. Counting them measures the tool, not you.
Measure instead: options killed, and the reason each one died. The judgment sits in what you rejected, not in what survived.
2. Speed ate the whole review
What the rubric says: hit the date.
Why it breaks: 45% of designers name faster turnaround times and 37% name higher expected output volume. Speed is easy to write down, so it crowds out every part of the job that is not.
Measure instead: rework. A first draft that gets rebuilt twice was never fast.
3. The job grew and the ladder did not
What the rubric says: the same levels as last year.
Why it breaks: 65% of designers say they are taking on more product or engineering work. Only 8% of leaders changed their performance metrics, and 4% changed pay structure. People can do that arithmetic.
Measure instead: whether the ladder names the new skill. If it does not, you are asking for that skill free.
4. Headcount fell and the targets held
What the rubric says: nothing. This one is a budget decision wearing a rubric.
Why it breaks: 19% report that their company cut headcount while expecting the same output or more. When output is the only measured thing, output is the only thing anyone squeezes.
Measure instead: surface area, meaning how many teams and products one designer now covers. That number explains the tiredness the output chart cannot.
5. The bottom rung went missing
What the rubric says: juniors grow by doing production work.
Why it breaks: Production work is exactly what went first. Meanwhile peer learning more than tripled, from 24% to 80%, while learning from leadership fell from 32% to 16%. The ladder is being rebuilt sideways by colleagues, not by companies.
Measure instead: reps at judgment. Give juniors a call to make and a critique to defend, early and often.
The swap, in one table
| What the review counts | What it misses | Better line |
|---|---|---|
| Screens shipped | The options you killed | Decisions recorded per project |
| Days to delivery | The rebuilds afterwards | Rework rate |
| Years at level | Skills the role just gained | New skills named on the ladder |
| Output per head | How much ground one person covers | Teams supported per designer |
| Tools adopted | Whether the work got better | Quality held at the new speed |
The audit
Open your own review document this week and count the lines.
How many measure something a prompt can produce? How many measure a decision you made and then defended? If the first number is bigger, your rubric is grading the tool and calling it you.
Then rewrite one line. Not the document, one line. Only 13% of leaders have moved, so being early is still on the table.
Get the PDF