How Many Shop Drawing Revisions Are Normal on Custom Millwork

What a Normal Revision Count Looks Like

Two rounds of shop drawing review is normal on a custom millwork package, and three is normal once radius work, historical matching or a full kitchen sits inside the scope. A first submittal goes out. The reviewing party marks it up once, in one consolidated response. A revised set comes back and gets stamped. That is the shape of a healthy job, and it holds whether the package is eight sheets of trim elevations or forty sheets covering cabinets, doors and beams.

Packages that run to five and six rounds almost never fail on draftsmanship. They fail because a decision that belonged to the design team was still open on the day the drawings were issued. A drafter cannot close an unselected appliance or an undecided countertop material. All he can do is draw the last thing he was told and wait for the markup that changes it.

ICD, the architectural millwork division of J. Gibson McIlvain, draws each custom item to the space it will occupy rather than to a catalog module, which front loads the questions. A shop drawing for a true custom cabinet run asks about scribe allowances, toe kick returns and the exact obstruction behind the sink wall before anything gets cut. That feels like more paperwork at the start. It is the reason the back end stays at two rounds. The ICD working process is built around resolving those questions on paper.

Revisions Versus Clarifications

Roughly half of what gets logged as a revision is a clarification, and separating the two keeps a project from looking troubled when it is running fine. A revision changes drawn geometry, material, joinery or finish. A clarification answers a question without moving a line. Logging both as revisions inflates the count and makes a normal job read like a problem job in a progress meeting.

There is a third category worth naming. A correction is the shop fixing its own error, a mislabeled section mark or a stile width that does not add up. Corrections should be rare and they should never consume a full review cycle. A responsible drafting group issues them as a marked sheet rather than reissuing the whole set and restarting the clock.

Scope changes are the expensive category and the one that quietly drives round four. Adding a bench to a window seat detail, swapping a flat panel door for a raised panel, or moving a beam centerline after the ceiling framing is up are all new work, not review comments. They deserve their own submittal rather than being folded into a revision of the original.

What Has to Be Settled Before Round One

Five decisions account for most extra rounds, and every one of them can be closed before a single sheet is issued. None require special expertise. They require someone with authority to answer, which is usually the harder part on a fast job.

Species and grade come first. Sapele, Iroko, Afrormosia, Teak and Utile are the working palette for milled and glued millwork, and they behave differently enough that a late swap redraws member thicknesses. Sapele runs around 1,410 lbf on the Janka scale and machines cleanly for painted and stained work alike, while White Oak sits near 1,360 lbf and Hard Maple near 1,450 lbf. The published property data for Sapele is worth reading beside the drawing rather than after it.

Verified dimensions come second. J. Gibson McIlvain supplies and mills, and it does not install or take field dimensions, so the numbers on the drawing come from the buyer's installer, fabricator or general contractor. A set drawn from design dimensions on a building that has already been framed generates a revision the moment someone holds a tape to it.

Finish intent is third and it gets skipped more than any other. Painted, stained, primed only or delivered raw changes reveal dimensions, the tolerance on a scribe and whether a joint can show at all. Hardware and appliance cut sheets are fourth, since a drawer slide length and a range width are drawing inputs rather than field decisions. Countertop material is fifth, since a heavy stone slab changes the cabinet construction underneath it.

Revision Rounds by Millwork Type

Review cycles are predictable by item type, and the items carrying real geometry or a historical match reliably take one round more than linear work. The table below is what a clean package tends to run when the five decisions above were answered before issue. Anything left open pushes each row one column to the right.

Notice what is missing from the driver column. Almost nothing in it is a drafting problem. The drivers are decisions, confirmations and outside approvals, which is why a shop that pushes hard for information early looks slow for a week and then finishes ahead of the one that started drawing immediately.

Why Radius Work Draws the Extra Round

Radius trimwork and radius millwork take a third round more often than any other category, and the reason is that a curve as designed and a curve as built are rarely the same curve. An elliptical opening laid out in CAD gets framed by hand. The finished radius drifts. Drywall thickness varies around the arch. A head casing drawn to the design radius stands off the wall near the springline, and nobody discovers it until the piece is on site with a crew standing under it.

The first review round on curved work should be spent confirming geometry and nothing else. Springline elevation, true radius at both jambs, whether the arch is a segment or an ellipse, and how the curve terminates into the straight run. Settle those and the profile discussion becomes simple. Reverse the order and the profile gets drawn twice.

Segment layout is the second common driver. A curved casing can be cut from laminated stock bent to a form or built from segments with joints falling in planned locations. Joint placement is a visual decision as much as a structural one, and a designer seeing those joints for the first time on a submittal will ask for them to move. That is a legitimate comment and it is also a full round.

Profile continuity matters more here than people expect. The same knife has to carry the straight casing and the curved head so the reveal does not jog at the transition. ICD draws the curved and straight members together on one sheet for exactly that reason, since approving them on separate sheets is how a mismatch survives review and shows up in a finished room.

Box Beams, Rafter Tails and Pergolas

Exterior and structural-looking items revise on joinery, moisture and length limits far more than on shape, and all three can be closed in the first submittal. A box beam looks like a simple three or four sided assembly on a drawing. The corner detail decides how it reads from the floor and how it behaves once it is up.

ICD builds box beams three ways and the choice belongs on the drawing, not in a phone call two weeks later. Butt joints give the most minimal profile, the fastest turnaround and the easiest on-site assembly, though they leave visible seams and visible end grain and they are the weakest of the three. Dado joints are strong and still assemble easily on site, but the seams and end grain still show. Lock miter joints show no visible seams and are the strongest, which is the usual answer on a stained beam at eye level.

Moisture content is the second driver and it is a specification line, not a preference. J. Gibson McIlvain mills interior millwork to an in-service moisture content near 6 to 8 percent and exterior work near 12 to 16 percent, so a beam drawn for a conditioned great room is not the same beam as one drawn for a covered porch. Moving the location after approval changes the milling target. The Forest Products Laboratory Wood Handbook carries the movement math behind those numbers.

Length limits catch pergola and rafter tail packages late. Modified woods top out around 16 ft, and most are sold on metric lengths just under 16 ft because of kiln size. Thermory and Abodo Vulcan both sit in that range. Thermally modified Ash is not available wider than 8 inches, so a wide modified member means Accoya or Abodo Vulcan instead. Accoya runs about 1,600 lbf with acetylation giving dramatically reduced shrinkage and swelling, and modified products are the only place a warranty exists at all. Solid unmodified wood carries none, since it is an organic material responding to its environment.

Cedar remains a sound default for pergolas and exposed tails, and the grade call belongs on the drawing. CVG, clear vertical grain, holds a crisp arris on a shaped tail and behaves predictably under a finish. STK, select tight knot, reads rustic and is entirely appropriate where the architecture wants that. Drawing one and delivering the other is a revision waiting to happen. Span and connection questions on a pergola frame are a code matter, and the American Wood Council standards are where the engineer of record goes for them.

Custom Doors, Gates and Shutters

Doors, gates and shutters revise on hardware, engineering and outside approval, three things that live outside the drafting room. Geometry on these items usually gets agreed quickly. What holds them up is a lockset nobody has selected or a historic commission that meets once a month.

Hardware drives door revisions almost single handedly. Stile widths have to carry the mortise. Rail heights have to clear the bottom of a vision panel. Swing direction, bevel and prep locations are all drawn, so a hardware schedule arriving after the first submittal guarantees a second. Custom doors and windows come off a drawing that already names the hardware rather than leaving a note to verify.

Gates split into two different problems. Pedestrian, garden and courtyard gates are craftsmanship, and they resolve in two rounds like any framed panel assembly. Large driveway swing gates need engineering before craftsmanship, because size and weight govern the frame, the joinery and what the posts have to carry. Plan a third round there and schedule around it instead of calling it a delay when it arrives.

Shutters carry the outside-approval problem. ICD builds louver and panel shutters and matches historical examples for both style and species, which is what a historic district requires when replacement has to be like for like. The review body, not the shop, sets that pace. Most local commissions work from the Secretary of the Interior's Standards for the Treatment of Historic Properties, and reading it before the first submittal prevents the round where a louver spacing gets rejected on principle.

Kitchen Cabinets and the Late Appliance Change

The most common cause of a third and fourth round on custom kitchen cabinets is an appliance or countertop decision made after the layout was approved. Move a range 2 inches and the flanking cabinets, the hood surround and the backsplash reveal all change. That is not a markup. That is a wall redrawn.

True custom cabinetry is engineered around how the space gets used rather than assembled from standard modules, which is what produces filler strips and mismatched face frames on a modular job. That approach absorbs odd conditions well. Storage can be worked around a waste stack or a mechanical chase that will not move. Specialty drawers can be drawn to a specific pan set rather than to a stock size. Cabinets and case work get drawn that way from the first sheet.

Countertop material belongs in the first submittal for a structural reason. Construction is tailored to the material being carried, and a thick stone slab asks more of the case than a thin engineered top does. Deciding it late changes the drawing beneath the counter, not just the edge detail at the front.

Construction standards cut rounds here. Naming an Architectural Woodwork Institute grade in the specification settles a long list of questions that would otherwise arrive as review comments, including joinery expectations, flatness tolerances and how a face frame meets a scribe. A specification that names a grade gets a submittal drawn to it.

Profiles, Knives and the Round That Never Has to Happen

A profile match is the one revision driver a shop can eliminate outright, and it comes down to whether the knife already exists. A drafter working from a photograph draws an approximation. The reviewer rejects it. That round repeats until somebody supplies a physical sample, a rubbing or a cut section of the original.

J. Gibson McIlvain holds a moulding profile library of thousands of ground knives and grinds new ones every week, so a profile already living in the library goes onto a drawing as a known section rather than an interpretation. When a profile is not in the library it gets ground to match, including difficult historical restoration profiles and contemporary sections no catalog carries.

"When somebody sends me a two inch offcut from a hundred year old house, that is the end of the profile conversation. We grind to it and the drawing is right the first time. When all I have is a photo taken from across the room, we are going to spend two rounds guessing at a bead."

Norm Moton, Director of Sales, J. Gibson McIlvain

Seventh generation and family run since 1798, J. Gibson McIlvain keeps knives and records on file, which matters on multi-phase work. A building getting an addition three years after the original trim was run needs the same section, not a close one. The record on file is what makes phase two a one round job instead of a rediscovery.

Finish Decisions Made Late

A finish schedule arriving after drawing approval reopens the drawing, because primed, painted, stained and raw are different products with different tolerances. Joint tolerance on a painted assembly is not joint tolerance on a stained one. Reveals change. Back priming requirements change. None of that is a coating question alone.

J. Gibson McIlvain primes in three levels, in both oil-based and water-based systems, and exterior primed trim carries a fungicide additive. Open-grained and oily species absorb a great deal of finish, so a primer coat evens out the eventual painted surface, and sanding or buffing after priming knocks back grain raising while keeping moulded detail crisp. Settling that at the drawing stage means the profile gets drawn with the finish film in mind.

Species choice for painted exterior trim belongs in the same conversation. Sapele is the better call than Genuine Mahogany for painted exterior work, since it holds paint better and comes in a wider range of sizes and longer lengths. Sapele is not milled 1x8 for exterior use, since that width is too wide to stay stable. Cupping in service is driven by installation and back ventilation rather than by species, which is a point for the installer rather than for the drawing. Building Science Corporation on moisture and materials covers the assembly side of it.

Quartersawn and rift stock moves less across its width than plainsawn, which is why it gets specified where a painted or panelized surface has to stay flat. Calling that out on the first submittal costs nothing. Discovering it after approval costs a round and sometimes costs the material.

How J. Gibson McIlvain Would Specify This

Write the submittal schedule for two rounds, budget a third for radius work, historical matching, large gates and full kitchens, and close the five open decisions before the first sheet is issued. That is the method. Everything past it is administration.

Name the species and the grade in the specification, using the millwork palette of Sapele, Iroko, Afrormosia, Teak or Utile for milled, glued and machined components. The tropical decking species, Ipe, botanically Tabebuia spp, along with Cumaru, Garapa, Jatoba and Red Balau, are extremely hard and carry high oil content that makes them difficult to glue, which is why a millwork shop reaches for the first group on a door, a beam or a cabinet even when the same project has a deck built from the second.

State the moisture target by location, 6 to 8 percent interior and 12 to 16 percent exterior, and state the finish level with the drawing rather than after it. Supply a physical sample for any profile that has to match something already standing. Have the installer confirm and supply field dimensions, since J. Gibson McIlvain mills and ships rather than installing or measuring in the field, and a drawing is only as good as the numbers it was built from.

The company runs a millwork shop on site at the lumber yard for linear trimwork, finished cladding and primed trim, with ICD, its architectural millwork division, off site for true custom work, so a package holding a straight casing run and a radius door surround moves through one order rather than two vendors trading blame at the review meeting. J. Gibson McIlvain stores finished millwork for inventory control and ships nationwide, regularly to California, which lets a phased job release in the sequence the site actually needs.

Send the drawings, the sample and the open questions to J. Gibson McIlvain at 800-638-9100 or through the millwork service page, and expect the first response to be a list of what still has to be decided. That list is the cheapest round of the project.

Frequently Asked Questions

How many rounds of shop drawing review should I put in the project schedule?

Schedule two rounds for a package of linear trimwork, box beams, rafter tails or standard custom doors, and three rounds where the scope includes radius trimwork, radius millwork, historical shutter or profile matching, large driveway swing gates, or a full custom kitchen. Build calendar time around the reviewer's turnaround as well, since a set that sits for three weeks on a desk costs more schedule than the drafting did. Packages running past four rounds are almost always carrying an unresolved decision rather than a drafting problem.

Does a question from the shop count as a revision?

No. A clarification answers a question without moving a drawn line, and it should be logged separately from a revision that changes geometry, material, joinery or finish. A correction is a third category, meaning the shop fixing its own error, and it should be issued as a marked sheet rather than as a full reissue that restarts the review clock. Keeping the three categories separate stops a normal job from looking like a troubled one in a progress meeting.

Why does radius work almost always take one more round than straight trim?

A curve as designed and a curve as built are rarely the same curve. Framing drifts, drywall thickness varies around an arch, and the springline can sit off the drawn elevation. The first round on curved work should confirm true radius at both jambs, springline, whether the opening is a segment or an ellipse, and where segment joints fall, since joint placement is a visual decision a designer will want to weigh in on. Confirm geometry first and the profile only gets drawn once.

What one thing cuts the most rounds off a profile match?

A physical sample. A two inch offcut, a plaster or pencil rubbing, or a cut section of the original ends the guessing immediately. J. Gibson McIlvain holds a moulding profile library of thousands of ground knives and grinds new knives every week, so a profile already in the library goes onto the drawing as a known section, and one that is not in the library gets ground to match, including difficult historical restoration profiles. A photograph taken from across a room is the single most reliable way to buy two extra rounds.

Can J. Gibson McIlvain verify the field dimensions so the drawings are right the first time?

No. J. Gibson McIlvain supplies and mills, and it does not install or take field dimensions. Verified dimensions come from the buyer's installer, fabricator or general contractor, and they need to reflect the building as built rather than as designed. Once those numbers are in hand the shop drawings get built from them, and the usual dimensional revision round disappears.

Does a late appliance or countertop change really need a new submittal?

Usually yes, because it is new work rather than a review comment. Moving a range even 2 inches changes the flanking cabinets, the hood surround and the backsplash reveal, and countertop material changes cabinet construction beneath the top since the case is built for the weight it carries. Treating it as its own submittal keeps the original approval intact and keeps the revision log honest.

Are shop drawings for a large driveway gate handled differently from a garden gate?

Yes. Pedestrian, garden and courtyard gates resolve in about two rounds like any framed panel assembly. Large driveway swing gates are an engineering question before they are a millwork question, since size and weight govern the frame, the joinery and the load the posts have to carry, so plan a third round and set the hardware, opening width and post condition before the first submittal.

Which species choices reduce revisions on a millwork package?

Naming a species from the millwork palette of Sapele, Iroko, Afrormosia, Teak or Utile at the start avoids the redraw that a late swap causes, since member thicknesses and machining allowances follow the species. For painted exterior trim, Sapele is the better call than Genuine Mahogany because it holds paint better and comes in a wider range of sizes and longer lengths, though it is not milled 1x8 for exterior use since that width is too wide to stay stable. Tropical decking species such as Ipe, Cumaru, Garapa, Jatoba and Red Balau are extremely hard and high in oil, which makes them difficult to glue, so a millwork shop reaches for the first group on doors, beams and cabinets.

Sources and Standards Referenced

• Wood Handbook, Wood as an Engineering Material, USDA Forest Products Laboratory
• Architectural Woodwork Institute
• Secretary of the Interior's Standards for the Treatment of Historic Properties, National Park Service
• American Wood Council Codes and Standards
• Moisture and Materials, Building Science Corporation
• Sapele, The Wood Database
• Accoya Modified Wood
• J. Gibson McIlvain Millwork Service
• ICD Our Working Process