Forty-seven comments and one Excel file

Transmittal 2 comes back with forty-seven comments. Twelve on the P&IDs, most on the narrative spec, three that reference "see previous submittal" without a page number. Project manager assigns disciplines, sets a return date. Mechanical knocks out their sheet in a week. Document control starts the comment log in Excel — because the comment log is always Excel.

Three weeks later you resubmit. Every marked-up PDF has a response. The customer opens Transmittal 3, finds comment 31 still open on a vessel datasheet nobody owned, plus a new comment on the heat and mass balance — a file they did not review on Transmittal 2 because it was not in the session folder.

Comment rounds fail quietly

The failure mode is not ignoring comments. It is resolving them file-by-file without a package view. Comments close in silos. One open item blocks commercial acceptance. The team discovers on Transmittal 4 that the closed item was closed on a superseded PDF while the released file still has the original wording.

• Comment log lives in one spreadsheet; markups live in Bluebeam; status lives in email.
• No single list ties comment → file revision → responsible discipline → verification.
• "Closed" means different things: fixed, won't fix, fixed in native file but not re-exported.
• Customer adds scope mid-round — new files appear without the same review rigor as the originals.

Forty-seven comments is manageable. Forty-seven comments across twelve file versions is a reconciliation project.

What a closed round actually requires

Before you call a comment round done, someone needs to answer: for each comment, show me the file revision where it was fixed, who verified, and what else changed in that same export batch. That is tedious work with folders and email. It is straightforward work with a structured review session where every response links to a versioned file and the release step will not complete with open items.

A comment log that survives handoff

• One register — comment ID, source file, page, text, owner, status, resolution revision.
• No discipline closes their own items without a verifier on critical specs.
• Re-export rules: if the native file changed, the PDF in the session must update before close.
• Scope the session explicitly — if HMB was out of scope on T2, say so; do not surprise the customer on T3.
• Ship the log with the transmittal, not as a separate attachment three days later.

Iterative submittals are how EPC and modular equipment work actually gets done. The tooling should match the rhythm — grouped reviews, visual diffs on resubmit, audit trail from comment open to release — instead of forcing each round through a new Bluebeam session and a fresh hunt through Drive. That is the gap between markup tools and document-set review. Not sexy. Just the difference between Transmittal 3 and Transmittal 5.