Submittal log management: stop reviews from stalling the job

A submittal log is the master register of every shop drawing, product data sheet, sample, and mock-up your project requires — what spec section demands it, who owes it, where it sits in review, and when it must be approved to protect procurement lead times. Managing it well is one of the highest-leverage things a PM can do, because under AIA A201 the contractor can't perform work requiring a submittal until that submittal is approved — so every week a submittal idles in someone's inbox is a week of schedule risk on the material behind it.

The gap between paper and reality is well documented. Specifications commonly allow on the order of 10–14 days for design-team review. But a San Diego State University case study of a 12-story, 220,000 sf project (Pestana & Alves, "Study of the Submittal Process Using Lean Construction Principles") found actual submittal lead times averaged about 32 days — and, counterintuitively, the GC's own distribution step took about three days longer than its initial review, while architect-only reviews ran about 12 days longer on average than reviews that also involved an engineer. The delay isn't always where you think it is.

What a working submittal log tracks

A log that actually manages the process — rather than just recording it — carries more than a number and a status. The fields below are the difference between a register and a control tool:

Two findings from that study are worth pinning to the wall. First, the slow step is not always the design team: the contractor's own distribution cycle exceeded its initial review by about three days, meaning the package sat inside the GC's process after the architect was done with it. Second, complexity didn't predict duration — shop drawing reviews actually ran about ten days shorter than product data reviews. Both point the same direction: cycle time is governed by queues and hand-offs, not by how hard the content is to review, which means it's governed by whatever your log makes visible.

Where the cycle actually leaks time

• Non-compliant first submissions. Wrong format, missing compliance statements, partial systems — rejected on a technicality after days in the queue. A two-step review (quick format check first, technical review second) stops a 10-day wait for a 15-minute rejection.
• The GC's own desk. The SDSU data put GC distribution slower than GC technical review. Internal hand-offs need ball-in-court tracking just as much as the architect's queue does.
• Long sub-tier chains. A sub-vendor's data passing through a vendor, a sub-sub, a sub, and the GC adds a delay opportunity at every hop. Make the prime sub responsible for issuing and managing its tier's submittals directly.
• Sequence errors. A sub-component submitted before its parent component forces reviewers to hold or reject it — burning a cycle for nothing.
• Resubmittal drift. Re-reviews should be faster than first reviews (the comments are known); without a stated goal — say, no more than two resubmittals — third and fourth cycles become normal.

Working the log week to week

A submittal log decays the moment it stops being worked. The teams that keep real control over the process tend to run the same simple cadence:

• A weekly ball-in-court review. Fifteen minutes, every week, same day: every open submittal, who has it, how long they've had it, and what the contract or spec allows. Items sitting past their window get chased that day — not at the OAC meeting two weeks later.
• A look-ahead tied to procurement. The question isn't "what's outstanding?" but "what must be approved in the next 30/60/90 days for material to land on time?" Long-lead equipment gets walked backwards from required-on-site dates, and those submit-by dates get enforced like pour dates.
• Resubmittals tracked as their own population. A package on its third revise-and-resubmit cycle is no longer a paperwork problem — it's a design conflict, a substitution fight, or a sub who hasn't read the spec. The log should make those visible instead of letting them blend into the queue.
• Status that matches the stamps. "Approved as noted" is not "approved," and building from an as-noted submittal without incorporating the notes is a rework generator. The log entry should carry the actual review outcome, not a simplified traffic light.

The cadence only works if updating the log is cheap. When confirming a status means hunting through transmittals and stamped PDFs, the log drifts from reality within a month — which is exactly the lookup problem worth automating.

Answering submittal questions from the documents — fast

Half of submittal-log churn is really document lookup: What does spec section 08 71 00 actually require us to submit? Was this product approved, approved-as-noted, or rejected — and what did the reviewer's notes say? Which revision of the shop drawing is current? Does this "or-equal" submission match the basis of design? Each one means someone digging through specs, transmittals, and stamped PDFs.

That lookup layer is what IntelMS handles. Email the question; it answers from your project documents with a citation to the exact spec paragraph or stamped submittal page — so the log gets updated from evidence, not memory. It will also flag when a question touches a superseded revision or a conflict between documents, and anything contractual — like whether a rejection entitles you to a time extension — gets escalated to a human instead of guessed at. Review decisions stay with the design team; your team just stops losing days finding what the documents already say.

Frequently asked