How to Read a DOB Queue: A Pillar Guide for Operators

Most operators treat the Department of Buildings queue the way they treat the weather: something that happens to them. That is a mistake. A DOB queue is a structured, observable system, and the operators who learn to read it consistently land permits four to eight weeks ahead of their less literate peers — on the same building, in the same jurisdiction, with the same drawings. This guide is the pillar reference we use internally when onboarding new principals onto a market.

We will cover what a DOB queue actually is, the five signals you can read from the outside, the per-jurisdiction variations that matter most, and the specific moves that get a file out of a slow queue and into a fast one. Where it helps, we will link out to the deeper jurisdiction guides — our New York City permitting guide, Chicago permitting guide, and Los Angeles permitting guide all reference this framework.

What a DOB queue actually is

A Department of Buildings plan-review queue is, structurally, three queues stacked on top of each other. First is the intake queue, where a clerk or junior examiner checks the submittal for completeness against the jurisdiction's filing checklist. Second is the assignment queue, where a complete submittal waits for an available examiner with the right discipline. Third is the active-review queue, which is the actual examiner backlog. Each of these has its own length, its own velocity, and its own failure modes.

When operators say 'the queue is slow,' they almost always mean the active-review queue. But in our cycle-time data, the assignment queue is the silent killer on 40 to 60 percent of slipped permits in tier-1 metros. A file that sits unassigned for three weeks is not in review — it is invisible. Reading the queue starts with knowing which of the three you are sitting in.

“Every Department of Buildings queue is a stack of decisions made by humans under workload pressure. If you can read the workload, you can read the date.”

Signal one: intake rejection rate

Every DOB publishes, or quietly tracks, an intake rejection rate. In NYC DOB, recent quarters have run between 35 and 45 percent. In Chicago, the published number is closer to 28 percent. In LADBS, the intake bounce rate on tenant improvements has been trending up toward 40 percent. These numbers tell you what fraction of submittals never reach an examiner on first attempt.

Your job is to land in the bottom decile of that distribution. The single largest lever is a per-AHJ packet template — same cover sheet, same drawing index, same response-to-comments grid that the intake clerk has seen a hundred times. We cover the templating discipline in The External PM Playbook for Multi-Site Rollouts. For glossary-level definitions of the intake artifacts themselves, see the permits glossary.

Signal two: examiner-to-application ratio

Every plan-review department has a finite number of examiners and a roughly known volume of incoming applications per quarter. When the ratio compresses — usually because examiners retire faster than the city replaces them — the assignment queue lengthens regardless of how clean your submittal is. NYC DOB lost roughly 12 percent of its plan-examiner pool between 2023 and 2025. LADBS has run with a 15 to 20 percent examiner vacancy rate for most of 2025. Chicago DOB has been comparatively stable.

You cannot fix the ratio, but you can read it. When the ratio is compressing, the move is to (a) avoid the in-house queue entirely via certified third-party reviewers where the jurisdiction permits it, (b) front-load pre-application meetings to compress the eventual active-review window, and (c) submit early in the week — Monday and Tuesday submittals get assigned faster in every queue we have measured.

Signal three: queue age at the head

Most jurisdictions publish or expose the date of the oldest unassigned file in each discipline. NYC DOB BIS and DOB NOW both make this visible if you know where to look. LADBS publishes weekly. Chicago DOB requires a phone call. The number is the most predictive single data point you can get on your own permit's near-term fate.

If the oldest unassigned file in your discipline is more than 21 days old, the assignment queue is the bottleneck and your file will sit. If it is under seven days, the active-review queue is the bottleneck and you should be focused on reducing comment volume, not on intake speed. Confusing these two leads to wasted effort.

Signal four: comment density and round count

Every jurisdiction has a typical comments-per-submittal density and a typical round count before approval. The published targets are almost always one to two rounds. The reality, across our 2025-2026 program data, is two to four rounds for tier-1 metros with a median of 18 to 32 comments on the first round and 4 to 9 on the second. Your file's comment density on round one tells you almost everything about its probable total cycle time.

If round one comes back with more than 40 comments, the file is structurally weak and you are looking at a four-to-five-round cycle. Stop arguing comments individually and rebuild the package. If round one comes back with fewer than 15, you are in a two-round cycle and the goal is a clean, fast response-to-comments turnaround. Read the comment density before you read the comments.

Signal five: examiner identity and history

DOB queues are not anonymous. Every file is assigned to a named examiner with a track record — a comment style, a known set of pet conditions, a typical turnaround pace. In jurisdictions where examiner identity is public (NYC DOB, LADBS, most of Texas), reading the assignment is a real signal. Some examiners run hot on egress. Some are slow but generate few resubmittals. Some are fast and consistent. Your response-to-comments strategy should be tuned to who is reading the file, not to an average.

This is one of the highest-leverage pieces of institutional memory a permitting partner brings. Commun-ET principals maintain per-examiner notes across every active market, and we update them every cycle. The 'named relationship' line on our Methodology page is downstream of exactly this discipline.

Per-jurisdiction variation that actually matters

NYC DOB runs two parallel systems (BIS and DOB NOW) with very different queue dynamics. DOB NOW assignment is generally faster but comment volume on novel uses can be higher. Misreading which system your work type belongs in is the most common single source of NYC queue confusion.

LADBS publishes by district. The Van Nuys queue and the downtown queue can be three weeks apart on the same work type. Filing in the right district — and choosing the district when you can — is the single largest schedule lever in the LA market.

Chicago DOB's self-certification track changes the queue equation entirely. For work that qualifies, self-cert bypasses the active-review queue for a fixed permit fee, and most operators undertarget it.

Texas markets (Austin, Houston, Dallas) generally accept third-party reviewers for structural and MEP, which lets you bypass the in-house queue when it is congested. Our Houston permitting guide and Austin permitting guide walk through the per-market mechanics.

What to do when the queue is working against you

First, escalate visibility, not volume. A polite weekly status request to the assignment desk is more effective than a stream of unanswered emails to the examiner. Second, exploit every available bypass: third-party reviewers, self-cert, expedited tracks, pre-application determinations. Each of these exists in most major jurisdictions and is undertapped. Third, parallel-track everything that can be parallel-tracked — utility coordination, fire review, health review — so the building-permit queue is not the entire critical path. We covered the parallel-track economics in the Wonder DMV multi-jurisdiction case study.

Fourth, time your submittals. Avoid the post-Labor Day surge in every major metro. Avoid the December slowdown. Submit when the queue is short and the examiners are rested. This is not a trick. It is queue literacy.

What this looks like inside the platform

PermitPilot ingests live queue telemetry from every jurisdiction it supports — assignment-queue age, examiner load, comment density rolling averages, and per-examiner historical patterns where the data is public. The agent layer (see Inside the Agent Control Center) uses these signals to predict your approval date and to flag when the queue conditions warrant a tactical move. The same telemetry is what we read on every weekly executive brief for active engagements.

If your team is running a portfolio of permits across two or more jurisdictions and you do not currently have queue telemetry, send us your portfolio. We will respond within one business day with a scoped read on what your real queue exposure looks like and where the fastest interventions are.