Why Bids on an Unfamiliar Building Come Back With Questions
By Michael Stern, JDS Development Group
A bid set for an unusual building goes out to a short list of contractors, and the first thing that comes back is a list of questions. Sometimes that list runs to several pages, and it arrives well before the date the numbers are due. Developers, construction managers and architects working on buildings with few precedents learn to read it as the first honest feedback the documents have ever had.
Reading it that way is possible only because competitive bidding has a formal question period built into it. That period exists for a reason worth understanding before the first answer gets written.
Why Bidding Acquired a Formal Question Period
Competitive bidding rests on a single assumption, which is that every firm pricing the work is pricing the same work. The assumption holds only if every firm is reading the same documents and reading them the same way. Nothing about a drawing set guarantees that on its own.
A drawing is a description of something that has not been built yet, and every description leaves something to the reader. The convention that grew up around that fact, long before any building I have worked on was drawn, was to set aside a defined period during which bidders ask and the design team answers.
The instrument that carries the answer is the addendum. An addendum is a written clarification or change sent to every bidder at once and folded into the documents, so that an answer given to one firm becomes part of what all of them are pricing. The reason it goes to everyone is the entire point of the mechanism, since one price can be compared to another only if both were built on the same information.
The sequence matters in the order it happens. Questions are collected, answers are drafted by the architect of record and the engineers, answers are issued as numbered addenda, and only then are prices due. Where that order breaks down, which is to say where an answer reaches one firm informally and never reaches the rest, the bids stop being comparable and the exercise has quietly lost its purpose. The industry bodies that maintain the conventions that organize a specification have spent decades on exactly this problem of making documents legible to people who did not draw them.
What a Contractor Is Actually Pricing
An estimator prices labor, material, equipment, the time the work will occupy, and the risk carried by everything the documents leave unsaid. On a conventional building that last category is small, because the firm has built the same detail many times and knows what it costs them to do it.
On a building with few precedents that repetition is unavailable. The estimator is looking at a condition nobody in the office has priced before, and the responsible thing to do is to price the worst defensible reading of the drawing rather than the intended one. An estimator who prices the intended reading and is wrong has given away the company.
This shows up first in the facade. The curtain wall on a tower with an unusual geometry is a long sequence of fabricated pieces, and the questions of who makes them, how they attach, and what happens where the geometry changes are all answered by shop drawings that do not exist yet during the bid. So the bidder prices the uncertainty instead.
The same logic runs through the ground. At The Brooklyn Tower, Brooklyn's first supertall, the new building rises from the landmark Dime Savings Bank, and a contractor pricing that work is pricing the care that working alongside an existing structure requires as much as the tower itself. The questions that come back on a site like that are mostly attempts to make the unknown smaller before the number becomes final.
Access is the other half of it, and it is priced the same way. At 111 West 57th Street, the tallest and most slender residential building in the Western Hemisphere, designed by SHoP Architects, the tower was built in the courtyard of the landmarked Steinway Hall. A bidder reading that set is working out how material reaches the work, where anything can be staged, and what a delivery looks like on a day when the street is doing something else. Those questions arrive early and they arrive from everyone, because no drawing in a bid set answers them completely.
What the Spread Between Bids Tells You
When prices come back close together, several firms read the same documents and reached similar conclusions about what is hard. That is useful information about the documents before it is information about anything else.
When the spread is wide, the firms read the same drawings differently, and at least one of them is carrying something the others have not seen. Chasing the low number in that situation usually means buying the reading that missed something.
The spread between bids measures how well the documents explain the difficulty, and the volume of questions measures the same thing a few weeks earlier.
A spread is also readable package by package rather than only in total. If the structural numbers agree and the facade numbers do not, the facade documents are where the work is owed, and the design team can put its remaining weeks there instead of spreading them evenly across a set that is mostly fine.
The volume of questions is the earlier signal and the cheaper one to act on. A heavy question list concentrated on one package is telling the design team precisely where the drawings are thin, which is worth more than any single opinion about the same drawings. On a building with few precedents that signal arrives alongside others, including what independent peer review of the structural design turns up, and the two tend to point at the same places.
Answering in Drawings Costs Time Nobody Budgets
My position is that a heavy question list should be answered in drawings rather than in prose. An addendum that adds a detail, a section and a note closes the question for every bidder in the same way. An addendum that explains the intent in a paragraph closes it differently for each of them, and every question left open at the award comes back later as work priced after the fact.
A competent estimator can argue the other side of that, and some do. Redrawing during a bid period resets the clock for every firm on the list, and firms that have already priced the original detail have to price it twice. There is a reasonable view that a clearly written clarification answered on schedule serves the project better than a better detail answered three weeks late, and on a package with a hard procurement date that view is often correct.
What I will not argue is that drawing it away always works. Some difficulty on an unprecedented building is real difficulty, and it prices as real difficulty no matter how good the documents are. A condition that has never been built before carries a number that reflects that, and the honest version of the exercise accepts the number rather than redrawing until it goes away. The bodies that publish structural engineering standards describe what a structure has to do, and nothing in the drawing of it removes the work of doing it for the first time.
The compromise I have accepted on this is slower bidding. Allowing a question period long enough to answer properly, and then a second short round after the addenda land, adds weeks to a schedule that never has weeks in it. Those weeks buy a set of prices that can be compared, and I have not found a cheaper way to buy that.
The people who read a bid set honestly are the construction manager and the cost consultant, and the people who answer it are the architect of record and the engineers of record. A developer who wants to know whether a bid is good should ask the construction manager what questions came in rather than asking what the number is, and the professional bodies behind construction management practice are a reasonable place to start on what that role actually covers.
A list of questions is the only free look you get at how well the building has been explained.
Michael Stern is the founder and chief executive officer of JDS Development Group.
