Estimating the Cost of Exotic Millwork


This article sets out some of the concepts and methods of estimating exotic millwork for installation in a variety of venues - from high-end residential, retail and hospitality (bars, restaurants and hotels), or office buildings such as law offices.

Brief Description
The millwork job described here was a residential project in one of the most affluent communities of San Diego, California. The points discussed also have some application to certain types of commercial work, especially in the hospitality industry or office environment.

Whether the job is design-build or design-bid-build, the owner(s) and  designer(s) will usually want a preview of the construction costs.  While some of these principles can be applied to a conceptual- or schematic-level estimate, I’ve tried to provide a theoretical and computational framework by which the estimator can narrow the fixity of a cost commensurate with at least a 90% design-development level set of plans and specifications. The reason for this is that a purely conceptual estimate can be done quickly, is based almost exclusively on historic data, and would not provide the challenge of an advanced design-development estimate, nor would it have the final commitment of price to the owner. The methods discussed here can be used to advance the estimate to a formal bid.

The set of plans for this type of project will have evolved from the architectural/design level to the shop level and be nearly complete, with only some details (such as molding call-outs and wood finishes) remaining unresolved. The plan set for the millwork/casework part of this job contained 136 pages. While these plans will remove much of the difficulty in determining cost based on owner-selected items and configurations, the challenges of pricing for general and special conditions -- as well as determining adequate labor and material sources -- will remain. Addressing the labor, material, equipment and risk factors that drive the cost of the millwork estimate continues to be crucial.

Project-Specific Factors to Consider in Takeoff and Pricing Small vs. Large Quantities

Even though a millwork project of this size and scope is usually thought of as a “large project,” breakpoints in molding pricing may not be available. The exotic nature of the project rests in its uniqueness and design as much as in scope. There may actually be more of a single type of molding in a small office than in a large office. The estimator is usually more concerned with pricing as it relates to length of materials and where to source them than quantity – especially in a job like this. The breakpoints occur when going over a certain length. Traveling to another city to inspect materials can quickly add cost as well.

The notion of large quantities also goes to the economic learning curve. This is distinguished from the trade learning curve as a journeyman gains more skill during his or her career. The economic learning curve relates to achieving an optimum speed of production with respect to repetitious work. There is almost nothing repetitious about this type of millwork, and it normally does not enter into the estimating process too much. An exception may be where there are very long runs of the same type of molding that can be done for many days without interruption, and where there may be some repetition of complex elements. However large a single millwork project may be, economies of scale usually also don’t apply in custom renovations as they would in tract work.

Special Conditions Affecting Millwork Installations

The following is a list of special conditions that the estimator should consider for any millwork job.

1. Vetting sub-tier contractors and suppliers. Proper vetting of sources is vital to a successful job. It is nearly impossible to quickly replace truly masterful installers on a moment’s notice, but having backups in case of emergency can be invaluable. Suppliers are another thing entirely, since they are fixed assets and well-known in advance, but you should always keep a list of specialty suppliers handy.

2. Proper backing. Proper backing and support of millwork systems is critical. The type of system and sometimes codes require certain types of supports for cabinets and paneling. This should be considered at design, during estimating and construction.

3. Coordination with every trade affecting the millwork. Since the millwork is a finished product, it must often contact other trades’ finished products. Coordinating this is essential. The job mentioned in this paper has product that is in contact with floors, Venetian plaster walls, faux finishes, metals and stone stair rails. All of this coordination takes time -- which must be calculated into the cost of the job. There is no substitute for experience here, and getting the installation crew in during some of the estimating can be very helpful.

4. Shifting schedules. All construction schedules change during the course of the job. Extra work orders affect the schedule—commonly a good thing. A surprise request to leave the job for a few days will also affect the schedule—potentially a bad thing. Millwork is consistently on the critical path due to its substantial scope and relationship to predecessor/successor tasks, and displaced tasks on the critical path can start affecting not only the overall labor cost due to demobilization/mobilization, but also the supply chain as it relates to lead times for certain important items and the possibility of needing to provide impromptu temporary storage. Who pays for this? The estimator will build this into the job, either by analyzing the historic probability of this happening (risk analysis) and adding that to the cost of the job, or by injecting contract language to cover its potential occurrence. More often than not, there are about two days of loss per whole crew for general contractor requests to millwork — to leave the job in order to insert another task in the schedule.

5. Productivity losses due to community CC&R regulations. Many high-end homes are in gated communities, with guards who open the gates right at the normal start of the construction work day or have regulations that effectively do the same thing. Without the roll-out time required to maximize the productivity of the work day, there may be less productivity. This is not about paying people for their time on the job; it is about how much work gets done in the day. Every job is different. Night-time work is almost never allowed except on commercial jobs. Some special venues such as mixed-use projects or those in occupied buildings carry this caveat as well.

6. Down time for expected trade work. One of the main types of work is anything that can affect accessibility to the job. Down time for exterior hardscape work is just one thing that can remove a millwork contractor from a job for a time. This should be planned for.

Labor, Material, Equipment, and Indirect Costs and Approach to Markups

Labor & Equipment Overview
Our crew was pre-selected based on past performance. Their rates were based on a one-hour unit and can be combined into a crew rate for some types of analyses. This rate is consistent with burdened wage rates in the area of the job at the time of construction. There were medical benefits added but no vacation or sick leave.

This job was local to the contractor; therefore, there were no travel costs to consider. Under other circumstances, these could include hotel, per diem, mileage, overtime to minimize time away, and possibly some city fees, since certain municipalities require licenses for out-of-town contractors.

The mechanical lift is billed out to reflect an hourly rate based on average use in the year with maintenance. Since it is in the middle of its life cycle, its rate is calculated on the basis of an aggressive depreciation curve.

The material costs are based on the quantity takeoffs and include all waste factors. This requires qualification. Most cut-off waste occurs at the shop since many pieces come to the job prefabricated. The cutoff waste consists of both waste from possibly incompatible lengths provided by the lumber company and snipe. The incidental waste at the job mainly relates to minor fitting of the already prefabricated pieces.

Since some travel and inspection of materials had to occur prior to purchase, there is a labor component in the materials themselves by the time they’re delivered to the job. This is aside from the shop labor required to manufacture the finished product. It is helpful to track this and keep it as a part of historical costs to measure it against future jobs. One can use this idea to simplify conceptual estimates in the future as jobs are reduced to unit time, unit area or whatever other measure might be useful. This becomes more meaningful as time goes on because a company usually settles in on a select group of vendors for similar components.

There were at least six lumber suppliers, two equipment suppliers and five hardware suppliers in this job. There were also three fabricators of specialty products for the foyer alone. For the lumber, some special components were outsourced and some were manufactured in house. All of the curved moldings were outsourced. These included moldings for the arched openings, curved base under the stairs and curved panel moldings for the stairwell itself, some of which followed a helical path. The curved panels and the curved jambs were all manufactured in house.

Some of the indirect costs include home office overhead and performance bonding. The contractor had to also include costs for special layout situations that used managerial skills. This included field measuring the curved panels, which involved calculating a circle through three points with which to compare with the architect’s plans and the framer’s product, as well as a field template to verify all of this. These panels were thick and fairly rigid: They needed to have minimal working to install and leave little residual stress, so they had to be as true as possible. These measurements were used for the handrail and curved molding manufacturing as well, and for the same reasons. Since the arrival of these products to the job had to keep with the schedule, the lead times had to be known in advance and strictly adhered to.

The factors that go into bidding for the desirability of a job are well-known, so I won’t discuss them here. However, there are several key components that go into finalizing an estimate of this sort. They are:

  • Replaceability of exotic materials damaged during manufacturing or installation as a risk event.
  •  Labor cost variations expressed as a factor in the estimate. The burdened labor base rate is a starting point. There should be a cell in the spreadsheet (or whatever software is being used) to adjust this according to the judgment of the estimator and upper management.
  •  The estimate should express as a cost factor various risk factors as a separate cell in the spreadsheet. They may not be all at the bottom near the “grand total.” While they will vary by job, some of these risk factors are weather, labor trouble, risk of injury (workers’ compensation rating) as well as past injuries, disposition of the owner or other key stakeholders as it relates to potential issues, interaction with other trades and the availability of key materials. Lastly, the cost to borrow money may have a small effect on the bottom line because, if many items are purchased on credit, there may be interest to figure in. Solvency and cash flow in the company itself can affect the labor cost as well if the intervals for job revenues require the contractor to use a line of credit to pay the installers.

Special Risk Considerations

Some risk considerations were mentioned in “Special Conditions.” Those were ordinary risks, but there are some special risks that should be considered for millwork jobs of this type.

  • The owner rejects some of the work. The owner may simply not like work that is installed per plans and specifications, but does not meet his or her expectations. Management of end customer expectations starts with the design phase. Whether the millwork contractor has a contractual relationship with the owner directly or the designer, the owner and the millwork contractor will interact at times during the job. Somehow, either the designer or millwork contractor (or both) have to set the customer’s expectations in such a way that. when the owner rejects some work, he or she understands that they are responsible for reworking that part. This happens often in high-end millwork.
  •  Unmet lead times. Lead times are very important, especially for long-lead items such as custom items fabricated specifically for the job. There is usually no excuse for significant delays in preset lead times. The estimator or project manager should work hard to absolutely ensure that the job materials come together as expected (and when expected). This starts with vetting the sources. This also may add cost to the job. The estimator or his or her representative should visit the facilities of companies that manufacture some of the job components and monitor progress.
  •  Failure to perform. There should be significant monetary damages for failure to perform, and the estimator should have a backup source for important items. The responsibility for delayed deliveries rests with the millwork contractor, but through contract language the owner should accept delays beyond the reasonable control of the millwork contractor. This would include wide-spread labor disputes and other similar factors.
  •  Special construction methods. This is also critical since a project, by definition, is a unique thing, and it sometimes requires unique methods and tools to accomplish. This job was no different. The foyer of the home had large, elevated and heavy panels that required special fastening methods to ensure that they would remain in place in case of a moderate seismic event. Even the glue that supplemented the metal fasteners had to be chosen so that it would not become brittle with time, thus ensuring that it would not fail. Some of the other curved elements required a special, custom-made fastening system that had to be custom designed.
  •  Special collaboration with suppliers. There are special metal finishes mentioned below. In the case of the butt hinges for the doors, the estimator had to make sure that the finisher understood that all hinge pairs had to remain together during the finishing process and be delivered to the job as originally sent from the manufacturer. Any inadvertent separation of the components of the butt hinges would have created a mismatch and caused improper door operation such as binding and squeaking. Not knowing this would have resulted in the rejection of the entire batch of hinges. This is not a short-lead item.
  •  Scope creep and gold plating. While the incidence of requests for gold plating the work was nearly non-existent on this job, it can happen. When the customer requests that one more little incidental item, it can seem a small thing to say, “OK”. However, collectively, they can quickly erode the profits of the job. The customer knows that he or she is spending a lot of money for exotic millwork, and sometimes that in itself creates a sense of entitlement: they’re king or queen for a short time, and they know it. This leads to a risk that can appear unexpectedly, and its management starts with the first meeting with the customer. Even if the estimator is not present at the initial meeting with the customer, he or she should become aware of the personality factors that may affect the job. There is no substitute for the ability to read people, and sometimes just knowing what a person does for a living can give insights into what to expect from them during the job.
  •  Self-performance or outsourced labor. This consideration may seem trivial, but while any sufficiently skilled labor force can successfully complete a high-end millwork job, sometimes it may be advantageous to use a sub-tier contractor for this. The reason is that the sub-tier contractor assumes more responsibility for the outcome, as well as responsibility for the use of valuable materials. This is effectively a risk-transfer device. The success of a sub-tier contractor is more assured if he or she is frequently used for this type of work.

Ratios and Analysis—Testing the Bid

Aside from counting square footages, lineal footages and adding up labor values for the job, the bid analysis mainly consists of comparing this estimate with the historic estimates of both won and lost jobs. The estimator should have basic benchmark data for installing panels, moldings and doors. These can be compared and adjusted by wood species, height on the wall and complexity, as in the case of multi-stage moldings and other factors. In project management, the “lessons learned” part of the project should be well-documented, and these can be invaluable in providing insight into where the lost time areas of a particular job occurred in the past.

Both the quantifiable information and the non-quantifiable information should come together and result in an appropriately adjusted price for the job. This price can be as high as a third of the cost of constructing the whole house, and can be well in excess of $1,000,000. Millwork is costly, and the larger the project, the more the cost. If, given the presence of cabinets in expected locations, ample special moldings and the panelizing of the entirety of at least two main rooms, the cost is less than about 25% of the total cost of construction, then the estimator should review the entire millwork job estimate.

Analysis is somewhat ongoing since the spreadsheets can be articulated in such a way that the information is exchanged between the layers of types of information. Linking information is not only important for the estimate summary, but additional sheets can be added to provide for increases to the company’s historic cost data. As a matter of estimating comfort, not only should others check the estimate, but the estimator should provide for internal verification of the estimate within the spreadsheet itself. This can take the form of checksums and intermediate order of magnitude calculations that can give a reasonable assurance of being on track.

Other Pertinent Information

The materials for a large millwork job must be sourced from multiple companies. This is not to build in unnecessary redundancy, but it is generally understood that there is no one-stop shop for wood products. One supplier may have one type of molding in a certain species, but not other types. If a custom molding is desired to match a catalog product, then it is sometimes necessary to travel to a supplier of raw wood products to make an inspection and a purchase of specified lengths of a species of wood to match the catalog product.

There can be many sources of paneling and molding for a single project. Book-matched and pattern-matched paneling has to be done carefully, since the size of the log determines the number of panels obtained from it, and that number varies by species and source. Paneling companies can provide guidelines and set expectations on the number of panels based on a particular species of wood. With the right kind of thought, the job can be built to accommodate a few errors in material usage, provided no one area uses all of one run of paneling. This project brought in wood from all over Southern California, and included rift white oak, red oak, mahogany, makore, chestnut, alder, white maple and sapele.

Special finishes are important to all millwork installations. In spite of the fact that millwork falls in the CSI Division 6 category, its jobs can include glazing, metals, storefronts and other specialty items. Clearly, it can work across several CSI Divisions. In this case, the customer chose hardware for style and in a finish that was not available from the catalog. This meant that all of the hinges and door locksets had to be sent out to be custom finished as oil-rubbed bronze.

Sample Plan and Profile View
Plans from Millwork Job

Figure 1: Floor plan view of Foyer. From “Smith Job Plan Set.” Copyright 2006 by Interior Wood of San Diego, Inc. Reprinted with permission.
Here we see a plan view of the entire foyer. The conceptual and descriptive elements of this paper are shown in Elevations 1/2, 2/3 and 3/4, and the sample estimate addresses parts of Elevation 1/2, the expansive flat wall of millwork.

Figure 2: Elevation 3/4 of Foyer. From “Smith Job Plan Set.” Copyright 2006 by Interior Wood of San Diego, Inc. Reprinted with permission.
This elevation shows the arched opening and the radiused wall. The panels surrounding the stairs, their bounding trim and the base molding are curved. Note the proportions of the arch whose shape appears in the profile below.

Figure 3: Arch profile. From “Smith Job Plan Set.” Copyright 2006 by Interior Wood of San Diego, Inc. Reprinted with permission.
This arch profile illustrates an arbitrary curve. This was arrived at by many design iterations with the designer and owner and was proportioned for the size and placement of the opening in the wall. It is only “mathematical” in the sense that it is piecewise continuous. It follows no equation.
Sample Millwork Detail Elevation and Section Detail

The following two drawings will provide information for our sample estimate.

Figure 4: Elevation 1/2. From “Smith Job Plan Set.” Copyright 2006 by Interior Wood of San Diego, Inc. Reprinted with permission.
This elevation shows the extents and dimensions of the estimated wall.

Sample Estimate—Takeoff and Pricing Sheets

Some of the ordinary outline headings such as described in the Standard Estimating Practice (p. 302) are: Item #, Item Description, Room #, Elevation #, Qty, EA / LF / SF, Height, Width, Length, #Cut-outs and Comments, and are found in most estimating forms. These find their way into various standard estimating forms and are adjusted by company, estimator or job particulars. Whichever way the information is organized, I suggest that the estimator use a layered approach in which the information is laid out in a fashion that goes from high-level to low-level in terms of granularity:

Item (here an assembly might be appropriate if the item is complex or repetitive)
Detailed Takeoffs and Supporting Information (often in a very raw form)

This sample estimate will cover the labor and materials for Elevation 1/2. The assembly idea works for this type of installation because experience has shown that the several installers that are needed to work on the large panels work together in their preparation and installation as a kind of mini fast-track project. Two of the four installers prepared the panel on a bench while the other two prepared the wall for receiving the panel, and all four were required to fit the panel. The bench was large enough to pre-assemble two panels at once.

Assembly #1: The Multi-Stage Midsection Band
The following spreadsheet shows the building of the estimating assembly for the multi-stage band at the wall midsection. The unit cost has been rounded to the nearest dollar.

Assembly #2: The Two-Stage Panels and Related Moldings
The following spreadsheet shows the building of the assembly for the multi-stage band at the wall midsection.

Final Estimate Summary for the Elevation
The following spreadsheet shows the estimate summary with markups for this portion of the job. Assumptions of moderate uniformity were made for illustrative purposes.

Some Concluding Thoughts
In the above estimate segments, there are mixed labor and materials in an effort to save space. Spreadsheet software can easily extract the specific data by using the SUMIF or SUMIFS functions and present it in any way imaginable. Constructing repetitive or complex portions of the job as assemblies is clearly not the only way to create an estimate. However, doing it this way will allow the estimator to provide a basis for improved intuition in using the historic records of the company. If one has the LF cost of 1-, 2-, 3- or multi-stage crown or other assemblies, then one has the ability to quickly create conceptual estimates in the future and feel confident in their presentation. This method allows for the development of an intuitive sense of how complexity relates to unit cost. This also applies to labor, materials and equipment if one culls each type of data from the spreadsheet and summarizes it in the company’s historic records. It can provide a check on the judgment factor in finalizing an estimate. Whatever the “feel” of the job is, judgment should never override hard numbers unless there is good reason to do so.

About the Author:
Jeffry N. Griffith, CPE resides in San Diego, California and is a member of ASPE Chapter 4. He is a Project Specialist and Estimator III with FEMA and Fluor Corporation and works in estimating for disaster recovery in any of the U.S. interests. Jeff holds a BS in Construction Management and is a licensed general contractor in the State of California. Additionally, he holds the PMP credential from the Project Management Institute and has won awards from the ASPE and the CMAA. Jeff also provides expert witness services to the legal profession in Southern California.

Jeff has over 20 years of project management experience, and his estimating experience goes back over 35 years to his foreman days where he supplied material and labor quantities to management for framing jobs. He is experienced in developing estimates in the CSI MasterFormat, using RSMeans cost data and is acquainted with UniFormat.
In respect to this paper, Jeff has estimated millwork and casework in a variety of venues, including corporate, hospitality, laboratory, military, education and residential. Jeff’s largest millwork/casework estimating job consisted of providing an estimate for all such in a six-story hotel renovation in downtown San Diego. This estimate included a restaurant, several bars, including a roof-top bar and cabanas, special custom casework for sleeping units and their surrounds for all suites, wet bar areas in rooms so equipped, paneling in some hallways and the restaurants, coffered ceilings, foyers, conference rooms and offices.

A previous version of this story appeared in American Society of Professional Estimators’ “Estimating Today”. It is reprinted with permission.

Tags: Costs, Estimating Construction

Leave A Reply