Door and Frame Takeoff Sequence for Commercial Bids
Getting the takeoff sequence right catches scope gaps that show up as costly change orders later.
Door and frame takeoff has a right order, and most estimators who miss their numbers on Division 8 scope got the order wrong, not the math. The sequence runs door schedule, then elevations, then floor plans, then specification. Reverse any two steps and the bid will look complete right up until submittals, when the gaps that the sequence exists to catch show up as change orders instead.
Division 8 scope runs wider than most people outside the trade assume. It covers doors, frames, and hardware, but also borrowed lights, sidelights, electrified hardware, and access control coordination that crosses into Divisions 26, 27, and 28. Every opening in a building carries its own assembly, a slab, a frame, a hardware set, specific preps, and a fire rating, and all of it has to reconcile before a number goes on the bid. The fire rating system alone adds real complexity. Commercial buildings use five standard ratings, 20, 45, 60, 90, and 180 minutes, and each one imposes its own requirements on door construction, frame prep, hinge type, closing device, latching hardware, and seal package. Miss a rating on one opening and the result is a failed inspection, not a rounding error on a spreadsheet.
Hardware set sprawl compounds the problem. Large commercial projects can carry more than 80 unique hardware sets, and each one needs pricing on its own terms. The cost of getting any of this wrong never shows up at bid time; it shows up downstream, in change orders, substitutions, failed submittals, and schedule delays. That lag is exactly what makes the mistake dangerous: it's invisible right up until it's expensive. None of this happens because the documents are missing information. The document set contains everything an estimator needs. It only works, though, if the documents get read together, in the right order, and most bid failures trace back to someone treating one document as sufficient on its own.
What the document set actually contains and why no single sheet tells the whole story
The door schedule is the master reference. It's formatted as a table, and every row represents one opening, tied to a unique door mark, something like D101, that links back to the floor plan. Columns typically cover door size, panel type, frame type, fire rating, hardware set reference, and detail markers for head, jamb, and sill conditions. That hardware set reference, something like "HW-3," is a pointer and nothing more; it tells an estimator where to look, not what's in the set. The content lives in the specification, and treating the schedule as a complete answer is the first mistake most people make.
Door elevations show configuration and prep. They reveal glazing cutouts, louvers, vision panels, and other features that the schedule row only summarizes. Where the schedule says "Type C frame," the elevation is what shows what that actually means for the opening's geometry.
Floor plans are the spatial check. Door tags on the plan sheets have to match the schedule, and every tag functions as a cross-reference, not a label to glance past. Plans catch what schedules miss. Openings that appear on a plan but never made it into the schedule are a real, recurring problem on commercial sets, and no amount of careful schedule-reading will surface that gap. Only the plan will. The wall type at each opening, shown on the plan, also determines frame selection.
The Division 08 71 00 specification is the controlling document for hardware. It contains the full hardware group listings: manufacturer, catalog number, finish, and function for every item in every set. On larger or more specialized projects, this might live in a separate hardware specification book, sometimes authored by a hardware consultant rather than the architect. And the spec can override the schedule. If the two conflict, the spec usually governs, but that has to be confirmed on each project, never assumed.
These four document types are built to be read against each other. Each one exposes gaps or raises questions that only one of the other three can answer.
Starting with the door schedule and why it comes first
The schedule establishes the universe of openings. Every step that follows is a verification against that universe, which is why nothing else in the sequence makes sense until the schedule pass is done. Start anywhere else, and there's no baseline to check the rest against.
The first pass is building the opening list. Every door mark gets logged in sequence, and a gap in the numbering, D104 missing between D103 and D105, is a flag, not something to explain away. For each row, capture door size, panel type, frame type, fire rating, and hardware set reference. Watch for "similar to" notations too; they always carry differences, and those differences need nailing down before moving on.
The schedule reveals scope that isn't obvious on a first read. A frame type reference points to a detail drawing somewhere else in the set, and the column value on its own means very little until that detail gets pulled. Fire rating entries need cross-checking against the assigned hardware set to confirm every component in that set is actually listed for the rating in question.
Version control matters before any of this gets committed to a count. Hardware schedules go through revisions on active projects, and working from a superseded version is one of the most common sources of wrong-material submittals later in construction. Messy schedules are the norm on real commercial bid sets, not the exception, and pretending otherwise is naive. Missing hardware set numbers, inconsistent frame type codes, rating conflicts, all expected. The schedule pass doubles as a quality audit, whether the estimator planned for that or not.
Reading elevations against the schedule to confirm what each opening actually is
Elevations are where the schedule's shorthand turns into an actual opening. A schedule entry might read "3070, HM, fire-rated, HW-12." The elevation shows the door's face, any lite cutouts, louvers or vision panels, and the frame profile surrounding it. Glazing kits, louvers, and vision panels get priced separately from the base door and frame, and they're easy to miss entirely if the elevations never get consulted, which is precisely how they end up missing from bids that looked thorough on paper.
Hardware prep coordination starts at this stage too. The elevation, and the door submittal that follows later, has to confirm the door slab arrives with the correct preps for whatever hardware set the schedule assigned it. Electrified hardware, electric strikes, electromagnetic holders, electrified hinges, requires prep at both door and frame that the elevation should show. If it doesn't, that gap belongs on the RFI log, not in the "figure it out later" pile.
Frame profile details need attention here as well. The frame type column points to a detail drawing, and that's the moment to pull it. Frame profile, material gauge, and anchor type all depend on the wall condition; a welded frame set into masonry and a knock-down frame set into metal studs are priced and supplied as entirely different products. Any opening where the elevation contradicts the schedule entry gets flagged, not resolved by guessing which document is right. Guessing is how the wrong frame gets ordered. Contradictions need clarification before the bid goes out, full stop.
Walking the floor plans to find what the schedule missed
The plan check isn't redundant work sitting on top of the schedule pass. It's the safety net underneath it, and skipping it because "the schedule already covers the openings" is the exact assumption that produces unpriced scope. Door tags on the plan sheets need to map one-to-one against schedule entries, and any plan tag without a matching schedule row is an unpriced opening sitting in plain sight. This isn't an edge case. Openings that show up on plan but never made it into the schedule are a documented, recurring failure mode on commercial sets.
Wall conditions read off the plan carry weight too. Masonry, metal stud, concrete, whatever the condition is, it directly controls the frame anchor system and often the frame profile itself. Interior partition schedules cross-reference here as well: a door sitting in a rated partition needs a frame and assembly matching that partition's rating, which can differ from the rating listed on the door itself.
Swing direction and clearance matter beyond the count. Door swings shown on plan affect hinge placement, closer arm selection, and whether the opening needs a wall stop or a floor stop. Plan reading isn't just a tally of openings; it's a configuration check. On large projects, floor plans run across dozens of sheets, and a systematic sheet-by-sheet sweep is what prevents the common failure of missing an entire wing, floor, or remote corner of the building. Any opening found on plan but absent from the schedule gets logged as a discrepancy, then either raised as an RFI or held as a contingency line. It should never quietly disappear from the count, which is what happens when this step gets rushed.
Entering the specification last, and why that order is deliberate
The spec defines what sits inside each hardware set. The steps before it define which sets apply to which openings, and that ordering isn't arbitrary; it's the whole argument for this sequence. Only after the opening list has been verified against plan and elevations does a hardware set reference like "HW-3" have a confirmed scope to price against. Pricing a hardware set before the opening count is confirmed means pricing a number nobody trusts yet, and that's the single most common shortcut estimators take when they're behind schedule on bid day.
Working through Division 08 71 00 means going group by group. Each hardware group lists every item: manufacturer, catalog number, finish, function. This is the actual line-item source for pricing, not the schedule. Grade requirements live here too. ANSI/BHMA Grade 1 is the standard for commercial occupancies and is required by most institutional specifications; Grade 1 cylindrical locks are tested to 800,000 cycles, and mortise locks to 500,000 cycles. The spec will name the grade, and often the test standard behind it. Fire-rated hardware groups need separate verification: every component in a fire-rated set has to be listed for the applicable rating, and the spec is where that check happens.
The spec generally controls over the schedule when the two disagree, but that has to be caught, never assumed by habit. Late addenda to hardware sets, changes issued close to bid day, get captured in the spec revision rather than the schedule. Reading the spec last means reading the most current, addendum-issued version of the document, which is exactly the point.
ADA and accessibility requirements surface in the spec as well. Hardware has to meet ADA operability requirements, including one-hand operation and compliant mounting heights. The spec calls this out for applicable openings, and the estimator has to price product that actually complies. On large or institutional jobs, the hardware specification might be a document authored entirely by a consultant, separate from the project manual, so confirming which document governs is its own necessary step before any pricing starts.
How skipping or reordering steps produces specific, predictable errors
Pricing from the spec before verifying the opening count is the single most common mistake in the sequence, and it's also the most avoidable. Hardware set quantities priced against an unverified opening list will be wrong every time the plan and schedule disagree, and they disagree more often than most estimators expect. Undercounting openings means underpricing the bid, and that error shows up as a loss on the job, not as a line-item discrepancy someone catches in review.
Skipping elevations and jumping straight from schedule to spec causes a different failure. Glazing kits, louvers, and specialty preps go unpriced entirely: items that sit right there on the elevation sheet but never appear in the schedule table. Hardware prep conflicts, a door arriving without the cutout its assigned hardware needs, surface during installation instead of at bid. By then it's too late to recover the cost.
Skipping the plan check causes its own version of the same problem. Openings present on plan but absent from the schedule get excluded from both the count and the bid. Wall conditions that determine frame type go unread, and the wrong frame profile gets priced without anyone noticing until submittal review, or later.
Working from a superseded schedule revision carries a quieter cost. Hardware sets that changed, additions, substitutions, deletions, get priced at their old configuration, and the contractor who wins the bid ends up installing to the current spec and absorbing the difference out of margin. None of these are estimator errors in the sense of a person being careless; they're process errors. A competent estimator following the wrong sequence produces the same wrong result every time, regardless of skill, which is the whole reason the sequence matters more than talent here.
Applying the sequence consistently across project sizes and document quality
Small projects, single-story retail buildouts, tenant improvements, carry fewer openings, but the same four document types still apply, and skipping steps because the job is small is a bad trade. Abbreviated schedules on small jobs are actually more likely to have gaps, because they get less scrutiny during design. The plan check matters proportionally more here too, since a single missed opening represents a larger share of a smaller total bid.
Large institutional projects, healthcare, K-12, higher education, shift the picture. Hardware specifications on these jobs are often written against an owner's master hardware standard rather than a project-specific spec, so the governing document at the end of the sequence might not be the 08 71 00 in the project manual at all. It's not unusual to see more than 80 unique hardware sets on a single institutional project, which makes the spec pass considerably longer and the hardware-group-to-opening mapping something that has to be done methodically, not from memory. BIM-originated schedules, exported out of Revit, tend to carry more reliable data during construction documents, but they still require the same cross-check against plans and elevations. Model errors propagate into a schedule just as easily as drafting errors do; the source changes, the risk doesn't.
Messy or incomplete bid sets need the same discipline regardless of project size. Missing hardware set numbers, clashing fire ratings, unlabeled frame types: each one is a flag requiring resolution, never an assumption filled in to keep moving. Logging every discrepancy during the takeoff sequence is what builds the RFI list; it's a byproduct of a disciplined process, not a separate task bolted on afterward. The sequence itself doesn't change based on document quality. It becomes more important when quality is lower, because the gaps and contradictions in a messy set are exactly what the sequence is designed to surface.
Where AI-assisted takeoff fits into this sequence and what it changes
The actual problem AI-assisted takeoff tools solve is a reconciliation problem, not a data-entry problem, and vendors who pitch it as faster typing are missing the point. Reading four document types at once and checking each against the other three is precisely the kind of task that breaks down under a manual, document-by-document workflow, because a person can only hold so many open questions in mind while flipping between a schedule, an elevation sheet, and a spec buried in a separate PDF.
An estimator working through a large commercial bid set sequentially, schedule first, then elevations, then plans, then spec, carries a growing list of unresolved flags for the entire duration of the takeoff: door marks with missing hardware references, frame types without a matched detail, openings seen on one sheet but not another. Software that cross-references a door schedule against plan tags and elevation callouts automatically doesn't replace the sequence described here; it enforces it. It checks for the same mismatches an experienced estimator is trained to look for, without the risk of one falling through during a long, multi-hundred-opening set. The sequence stays the same either way. What changes is how much of the cross-checking gets done by a machine before a human has to make the judgment call on a genuine conflict.
