A hiring rubric is good only if it still predicts performance a year after the hire.
Most rubrics I have inherited grade for portfolio polish, communication clarity, and culture fit. Twelve months in, they are noise. The polished portfolio came from a team that did the polishing. The clear communicator went quiet under load. The culture-fit signal turned out to be class signal in a costume.
Three criteria have held up for me across appraisal calibrations and the next round of hires I made off the back of them.
First, can the candidate change their mind in the room. Hand them a critique on a piece of work they are proud of. The wrong answers are defensive (“the constraint was X”) and over-eager (“you are completely right”). The right answer is a clean updated position thirty seconds later. Most senior portfolios pass screen one. This filter halves the field.
Second, can they name what they did not do. Every shipped product has a version that was cut. The senior designer can describe the version they argued against, why, and what they would resurrect if the constraint lifted. Juniors describe what was shipped. Mid-level designers describe what they wished was shipped. The senior describes the decision tree.
Third, can they hold a strong opinion without owning it. Junior designers either hold no opinion or hold one so tightly it becomes the conversation. The senior designer offers a position, marks the confidence level, and moves on if the room disagrees. The opinion is in service of the work, not their identity inside it.
The counter-view is that any rubric leaves out the candidates who interview poorly but perform brilliantly. Mine has missed at least one of those. The mitigation is a take-home piece that lets the work do the talking, scored against the same three criteria.
The shift is to stop hiring people who are good at being interviewed and start hiring people who would be useful in a calibration meeting.