Phoenix DSD makes plan-check correction lists publicly accessible after a permit decision. I pulled 500 of them over a Saturday and a Sunday — random sample of commercial TI and residential additions from the last 14 months. Coded every comment into a category in a Google Sheet.
Twelve specific mistakes account for over 80% of all rejection comments. Most plan-checkers in Phoenix DSD are seeing the same handful of issues, week after week, regardless of which firm submitted the set.
Here they are, ranked by frequency. If your set has any of these, plan-check will catch it.
The 12 mistakes, in order
- 1. Code-analysis sheet missing required tables (occupancy load, allowable area, egress count). 18% of all comments. Reviewers want to see the math, not just the conclusions.
- 2. Accessible-route arrows missing or terminating short of the accessible feature. 11%. Especially common around restrooms and primary entries.
- 3. Fire-rating callouts on assembly schedules without UL listing references. 9%. Reviewers want the UL number, not just "1-hour rated."
- 4. Egress-illumination plan missing or incomplete. 8%. Especially around exterior egress doors and hardware.
- 5. Door schedule without hardware groups specified. 7%. Hardware groups have to be referenced explicitly, even if simple.
- 6. IECC compliance documentation absent or incomplete. 6%. Performance-path projects need REScheck or COMcheck output included in the set.
- 7. Mechanical scope on the architectural set without coordination with the M sheets. 5%. RCP and HVAC plan-by-plan conflict checks expected.
- 8. Missing change-of-use review where occupancy is shifting. 4%. IBC chapter 3 evaluation must be on the cover sheet.
- 9. Wall-section details inconsistent with the floor plan. 4%. Materials called out on the section that don't match the floor-plan key notes.
- 10. Site plan not showing all required setbacks dimensioned. 3%. Especially side and rear setbacks on residential projects.
- 11. Title-block information incomplete (designer of record, owner contact, address). 3%. Sounds trivial; it isn't — title-block deficiencies are an automatic kickback at intake.
- 12. Smoke-detector / CO-detector locations not shown on the residential plan. 2%. R-3 occupancies require these on the floor plan, not just the electrical sheet.
What's interesting about the list
Several things, actually.
First — none of these are hard. Each one is a cover-sheet checklist item or a standard drawing convention. They're just things that get skipped under deadline pressure or by drafters who haven't filed at Phoenix DSD enough to know what plan-checkers always ask for.
Second — the ranking is consistent across project type. Commercial TI and residential addition sets get rejected for the same things at roughly the same frequencies. Code-analysis-sheet completeness is the #1 issue everywhere.
Third — these are the same mistakes I see in plan-check comments from other AHJs. Phoenix DSD just publishes them more readily than most. The list above is essentially a universal pre-submission checklist for any commercial or residential drafting project in the United States.
What to do with this
If you're a homeowner reviewing a permit set before your contractor submits it, check the 12 items above. You don't need to be an architect to spot most of them. If the code-analysis sheet doesn't have an occupancy table, ask why. If accessibility routes aren't drawn with arrows, ask why. The questions catch about 80% of the corrections that would otherwise come back from plan-check.
If you're a drafter or designer, build a pre-submission checklist around these 12 items. Run every set through it before you file. We do; it's the reason our first-review pass rate at Phoenix DSD is over 80%.
If you're a developer or GC trying to evaluate drafting firms, ask what their first-review pass rate is. Anyone over 70% is using a checklist roughly like this. Anyone under 50% is, statistically, making the mistakes above repeatedly. The difference shows up in your project schedule.
Plan-check rejection isn't random. The same 12 mistakes account for most of it, in Phoenix and elsewhere. Build a checklist; run every set through it before submitting; pass rate goes up.
I''m happy to share the spreadsheet with the 500-comment dataset to anyone who wants to see the underlying coding. Send an email and I''ll forward.