The 12 reasons submittal packages actually get bounced.

What the review stamp actually means

A package does not come back as a simple yes or no. It comes back with a review action, and the wording decides whether you can order material this week or wait out another full turnaround. The exact labels vary by design team, and they resolve to five outcomes.

Approved as noted is the outcome worth chasing on a first pass. It keeps your schedule intact while still letting the reviewer correct you, and it is the practical difference between a package that was close and one that was wrong.

What a bounce actually costs

The rejection itself is not the expense. The weeks behind it are, and they stack in a specific order.

• The review turnaround, again. Whatever the design team took the first time, they take again. That number belongs to them and you cannot compress it.
• The lead-time clock, still not started. Long-lead material counts its lead time from the approval stamp. A rejection does not pause that clock, it delays the moment it starts.
• Crew out of sequence. Work that was scheduled against a delivery date now has to be moved, and moved work rarely moves cheaply.
• The PM hours, spent twice. Compiling costs days the first time. A resubmittal does not cost the full amount again, and it is rarely the small correction it looks like either.
• Reviewer attention, permanently. An engineer who found one unchecked line reads your next package looking for a second.

Industry figures put the first-round rejection rate at roughly 30 to 40 percent, which means a sub running six packages a year should expect two of them to come back. Each one costs weeks: another full review turnaround, and a lead-time clock that waits on the approval.

The order to build in, so nothing is left to find

The 12 checks above are what a reviewer looks for. Building in an order that answers them in advance is faster than auditing a finished package against them.

• Start in Division 01. Submittal procedures and substitution procedure apply across every section, and a procedural miss bounces a technically perfect package.
• Then Part 1 of each section, for what has to be sent and in what form. Product data, shop drawings and samples are separate obligations.
• Build the product list from the drawings, not from memory. Schedules, then the legend, then keyed notes, then the rough-in that no schedule names.
• Check Part 2 before sourcing anything. The accepted-manufacturers clause decides which candidates are even worth pricing.
• Mark every multi-configuration sheet as you add it. Marking at the end is how sheets get missed.
• Flag conflicts rather than resolving them. A spec-versus-schedule disagreement is an RFI, and silently picking a side transfers it to you.
• Count the package against the schedules last. Completeness is checked before quality, so it should be verified last and deliberately.

The printable version of that build order is The Package Pre-Flight (BM-002), which comes with the checklist below. For the underlying structure, see what a spec section actually asks you to submit.

Common questions

What is the difference between rejected and revise and resubmit?

Revise and resubmit means the submission had problems. Rejected means the product did. A revise generally means fixing what you sent about a product that is still acceptable; a rejection usually means finding a different product, because the one proposed is not permitted by the spec.

Can I order material against an approved as noted stamp?

Normally yes, provided you order what the marks say rather than what you originally proposed. That is the entire point of the distinction, and it is also where subs get caught: the marks are part of the approval, and ordering the unmarked version is ordering an unapproved product.

How long does a submittal review take?

The contract usually specifies it, commonly ten to fifteen working days, and the specified number is a floor rather than a promise. That turnaround belongs to the design team, which is why the first pass matters more than the speed of your own compilation.

Does a partial package get partial approval?

No. Completeness is checked before quality, so a package covering 84 of 91 scheduled products comes back whole. The 84 that were correct get no credit and get reviewed again from the start.

Who is responsible when the spec and the drawings disagree?

The design team, if you flag it. You, if you do not. A conflict raised as an RFI before submission is theirs to resolve; the same conflict resolved silently inside your package becomes your decision and your liability.

Do I need a substitution request for an equal product?

If the spec says so, yes, and the quality of the substitute is not the question. Division 01 25 00 usually carries the procedure and a deadline measured from bid or award. Miss the form or the deadline and the product is not judged at all.

Why do reviewers reject on formatting?

Because the review is timed. A reviewer with ten minutes and no table of contents, no spec-section order and files named scan_004.pdf cannot verify the package inside the time available, so returning it costs them less than reading it.

What is the fastest way to lower our rejection rate?

Check the accepted-manufacturers clause on every product before pricing anything, and mark every multi-configuration product data sheet. Those two account for a large share of first-round bounces and both are cheap to fix.