How we run design reviews internally (our actual process)
A transparent look at our internal design review process — the templates, the cadence, and the hard lessons we learned along the way.
Design reviews are one of those things every team says they do, but few do well. For the first six months of AXI, our reviews were inconsistent — sometimes too nitpicky, sometimes rubber stamps. It took us a while to find a process that actually improves the work without slowing us down.
Async Critiques
We now run two types of reviews: async critiques and live sessions. Async critiques happen for every deliverable before it goes to the client. The designer posts the work in a dedicated Slack channel with context — the brief, the constraints, and any specific areas where they want feedback. Team members leave structured comments within 4 hours.
The structure is important. Every piece of feedback must include what level it is: L1 (must fix — accessibility issues, brand violations, factual errors), L2 (should fix — usability concerns, visual inconsistencies), or L3 (could improve — subjective polish suggestions). This prevents bikeshedding and keeps reviews focused on what matters.
Live Sessions
Live sessions happen once a week for 45 minutes. We pick 2-3 projects that are at a critical decision point — usually early concepts or near-final work. The presenting designer walks through their thinking, not just the output. We found that understanding the rationale behind decisions leads to much better feedback than just reacting to visuals.
The hardest lesson we learned: the person giving feedback is responsible for being constructive, but the person receiving feedback is responsible for not taking it personally. We explicitly talk about this during onboarding. Great work comes from honest critique, and honest critique requires psychological safety.
Share this article
