Before I moved into product work at FOBECON, I spent years as a practising quantity surveyor. Most of that time involved reviewing BOQs produced by other estimators before they went into tender submission. What I found was not random error. The same mistakes appeared across projects and firms with a regularity that suggested something systemic. The errors were not failures of arithmetic. They were failures of process, specifically the informal processes that estimating teams develop when working from PDF plans rather than structured data.
These are the five that came up most often, and why they are harder to catch than they look.
Mistake 1: Treating the door schedule as the count
Door schedules appear in most commercial drawing sets as a table listing each door type, leaf size, frame, hardware, and fire rating. Estimators frequently use this schedule as the definitive count of doors in the project. This is a mistake, and a common one.
The door schedule is a design-intent document produced by the architect. It lists the doors that should be in the project as designed. The floor plans, particularly in later-stage or as-built drawings, often contain additions, substitutions, or unlisted elements that did not get back-propagated to the schedule. A door added to a corridor during a design coordination meeting may appear on the floor plan before it appears in the revised schedule, or may never make it to the schedule at all if the revision was issued informally.
The count that matters for takeoff purposes is the count on the floor plans, cross-checked against the schedule. The schedule confirms type and specification; it does not confirm quantity. Estimators who use the schedule alone typically miss between 3 and 8 percent of the actual door count on a complex fit-out project. On a 200-door commercial building, that is 6 to 16 doors, each of which has a supply, frame, hardware, and installation cost attached to it.
Mistake 2: Measuring partitions without checking fire zone boundaries
Partition takeoff in most estimating spreadsheets has two or three categories: standard partition, high partition (above false ceiling), and acoustic partition. The classification determines the unit rate. What is frequently missing from this classification is a fourth category that has materially different cost implications: fire-rated partition.
Fire zone boundaries are shown on dedicated fire strategy drawings, not on general floor plans. Estimators working only from architectural floor plans and a partition schedule will count the linear metres correctly but apply the wrong rate to every partition on a fire-zone boundary. Fire-rated wall assemblies, particularly those requiring 60 or 120-minute ratings with specific stud spacing, intumescent strip details, and tested board specifications, cost substantially more per linear metre than standard partitions.
The fire strategy drawing is not always included in the standard tender drawing set. Estimators sometimes have to request it separately or assume from the specification that certain areas will require fire rating. The safest practice is to request fire strategy drawings as a standard part of the tender query, and to annotate the partition count with fire zone information before applying rates. In practice, this step is often skipped under deadline pressure.
Mistake 3: Counting from a single drawing type
A floor plan gives you a top-down view of a building. It shows horizontal extent well. It shows vertical elements, column positions and stair cores, but it shows them as plan projections and not as three-dimensional realities. Estimators who count structural elements exclusively from floor plans miss the scope implied by sections and elevations.
A common example: external cladding area measured from a floor plan perimeter produces an incorrect figure whenever the building has setbacks, overhangs, or varying floor-to-floor heights that are visible only on elevations. An estimator counting from the ground-floor plan perimeter multiplied by the number of floors will overcount in areas where upper floors step back, and will miss cladding on roof parapets and plant-screen structures that appear only on elevations and roof plans.
The same principle applies to vertical structural elements. A concrete column shown as a square on a floor plan at 1:100 has the same visual weight as a column at 1:50. The actual cross-section, and therefore the concrete volume and formwork quantity, requires the section drawing to determine accurately. Column schedules are more reliable for volume takeoff than floor plans, but only if the schedule exists in the drawing set provided for tender.
Mistake 4: Not recording the revision against which the count was made
This is an administrative error, but its consequences are significant. Drawing sets issued for tender are typically revision C or D at the time of bid. By the time a contract is awarded and construction mobilisation begins, the design may have moved to revision F or G. The BOQ used for tendering is tied to a specific revision. If the revision is not recorded in the BOQ header, nobody can efficiently audit the count against the drawings used to produce it.
When a variation arises during construction, the contract administrator asks the contractor to substantiate the variation claim. If the original BOQ count was made against revision C drawings and the current issue is revision G, the contractor needs to compare the two to isolate the scope change. If the original count does not specify the revision, this comparison requires reconstructing the count from scratch. That is expensive, and it weakens the contractor's position in variation negotiation.
The fix is simple: every BOQ should record the drawing number, title, revision, and date of issue for each drawing used in the count. In practice this is often skipped because it adds time to the counting phase. The time cost of the fix is roughly 20 minutes per project. The cost of not doing it during a contested variation can be measured in days.
Mistake 5: Applying a wastage allowance without checking the material type
Standard practice in quantity takeoff is to apply a wastage factor to material quantities before pricing. Typical allowances range from 5 percent for sheet goods with low wastage to 15 or 20 percent for irregular tiles or stone with high cutting requirements. Most estimating teams apply these factors from memory or from a standard table that has not been updated since the team leader learned the trade.
The problem is that modern construction uses a growing range of material configurations that do not behave like the standard cases. Large-format porcelain tiles (typically above 600x600mm) have significantly higher cutting waste on rooms with non-standard dimensions because fewer tiles span the room before a cut is required. Engineered timber flooring installed at an angle incurs higher waste than installation along the room's primary axis. Acoustic insulation batts in non-standard stud spacings require cutting that can push waste from 5 to 12 percent.
Applying a default 10 percent wastage to large-format tiles in a fitout that is primarily composed of irregular spaces, as is typical in a hospitality project, will underprice the material budget. The estimator who knows the project type, room geometry, and specified materials can apply the correct factor. The estimator working quickly from a specification schedule without cross-referencing the floor plan geometry will apply the default.
The pattern behind the mistakes
Looking at these five together, the common thread is cross-referencing. The schedule against the plan. The fire strategy against the partition count. The elevation against the floor plan measurement. The drawing revision against the BOQ header. The room geometry against the wastage factor table.
Each of these cross-references requires switching between documents, holding a mental model of what each document says, and reconciling discrepancies. Under time pressure, estimators skip the switches and rely on what is in front of them. The errors that result are not the result of incompetence. They are the result of a process that makes cross-referencing expensive and easy to skip.
We are not suggesting that eliminating the need for cross-referencing is possible or desirable. A quantity surveyor who checks their count against multiple drawing sources is doing their job correctly. The question is whether the mechanical part of that checking, verifying that a count made from a plan is consistent with a count made from a schedule, has to be done by hand every time.