Bid a job when two things are true: you can put a named superintendent and crew on it for the months it will actually run, and the drawings and specs define the scope well enough to price without guessing. Whether you can win it is not one of the tests. A job you win but cannot staff costs more than a bid you never sent — it loses money on itself and on every job it borrows people from.
Winning is the wrong test
Most advice on the bid no bid decision construction firms make starts with win probability. Relationship with the owner. Number of bidders. Whether you have worked with this architect before. Those factors tell you whether the job will come back signed. They say nothing about whether you will make money once it does.
A builder running 5 to 40 jobs is rarely hurt by the bids it lost. The cost of a lost bid is bounded — the takeoff hours, the sub coordination, an afternoon of pricing. The bids that hurt are the ones that won, started in a month the company was already full, and ran with a superintendent split across two sites. That cost is not bounded, and it does not show up on the job that caused it.
Test one: can you staff the months it actually runs
The question is not whether you have capacity today. It is whether you have a superintendent, a lead carpenter and the pace-setting trades free during the weeks this job will be in the ground. Those are rarely the weeks on the invitation. Permits run long. Financing closes late. The architect is still issuing addenda. A job bid for a spring start often breaks ground in summer, right on top of work you signed in the meantime.
Run the check before the full takeoff starts. It takes a rough estimate and a calendar, not a finished number.
- Break the job into phases by week from a rough estimate — sitework, structure, rough-ins, close-in, finishes. A single total duration hides the stretch where the job needs the most people.
- Set it on the calendar at the owner's stated start date.
- Set it again at a later start — the slip you have actually seen on similar jobs with this owner type, not a hopeful one.
- Lay both versions over the jobs already under contract and the pending bids you expect to win.
- For every overlapping month, write the name of the superintendent and the lead. A name, not 'we will hire someone'.
- Check the two or three trades that set the pace — usually framing, mechanical and drywall. If your framer is booked through the fall, the job is booked through the fall.
If any month comes up without a name in it, the job fails the test. The office counts too. Every job adds RFIs, change orders, pay applications and lien waivers. If the person who handles them is already behind, a new job makes every existing job slower to bill.
A superintendent hired after the contract is signed learns your systems, your subs and your paperwork on the owner's job. The alternative — splitting an existing superintendent across two sites — slows both. Either way the job runs longer than you priced it, and general conditions are a weekly cost. Every added week of trailer, supervision, temporary utilities and insurance comes straight out of margin.
Test two: is the scope defined enough to price
The second test is about the documents, not the company. Some bid sets are complete enough to take off and price line by line. Others are a design intent with a bid date attached. Signs the scope is not ready to carry a lump sum:
- Mechanical, electrical or plumbing shown as 'design-build by contractor' with no engineering behind it
- No geotechnical report on a job with foundations or significant sitework
- Finish schedules blank, or specs that name a product category but no product
- Owner-furnished items listed with no description of what they are or who installs them
- Allowances covering major systems rather than finishes
- Addenda still being issued inside the last week before bids are due
Any one of these can be handled. The response is to match what you send to what the documents support. Price the defined scope. Carry the undefined parts as stated allowances, unit prices or written exclusions. If the documents only support a range, what you can honestly send is an estimate, not a bid — the difference is laid out in estimate vs quote vs bid. If the gaps are too large to qualify around, decline.
What does not work is a lump sum over undefined scope with contingency spread on top. Contingency is for named risks inside a defined scope. Used to cover scope nobody drew, it either gets cut in negotiation or prices you out against a bidder who guessed low — and the bidder who guessed low is the one who wins.
Why a job you can't staff costs more than a bid you never sent
The comparison is easier when the outcomes sit side by side. Four things can happen once a bid invitation arrives:
- Decline early: costs the hours spent reading the documents and one phone call. Relationship risk is low if you say why.
- Bid and lose: costs the full takeoff, sub coordination and pricing time. Returns current sub pricing and a read on the market.
- Win and staff it: earns the margin you priced, give or take the usual change orders.
- Win and can't staff it: earns the priced margin minus extended general conditions on this job, minus the slip on every job it borrowed people from, minus the owner's willingness to call you next time.
The fourth outcome is the expensive one because the loss is spread out. The job itself may finish close to budget on paper, because its overruns land as weeks rather than line items. The job the superintendent was pulled from slips its punch list. Retainage on that job is held longer. Its owner gets a slower response to a warranty call. None of that gets coded back to the job that caused it.
Overhead makes it worse. Your construction overhead rate assumes a volume of work moving through the company at a certain pace. A job that drags across extra months absorbs overhead it was never priced to carry, and the office time spent managing the slip is time not spent billing other work.
What does not belong in the decision
These reasons come up in every bid meeting. None of them answers either test.
- 'We need backlog' — backlog you cannot build is a liability with a start date.
- 'The owner asked us personally' — a decline with a reason keeps you on the list. A late, short-staffed job takes you off it.
- 'We are due to win one' — win rate is a lagging number. It does not change what the next job costs to run.
- 'It is a big job' — revenue is not margin, and a large job that runs long consumes more supervision than two small ones.
- 'We can sharpen the number to get in' — buying a job you cannot staff stacks a thin margin on top of a schedule overrun.
Running both tests in one short review
The review works best before anyone starts counting. Twenty minutes with the bid set, the calendar and whoever assigns superintendents is enough to decide most invitations.
- Read the sheet index, spec table of contents and allowance schedule. Mark every scope gap from the list above.
- Build the rough phase durations and lay them over current commitments at two start dates.
- Write a name in every overlapping month for superintendent and lead.
- Decide one of three things: bid, bid with stated qualifications or decline.
- Write the decision and the reason in one line. Revisit declined jobs each quarter to see whether the reasons held.
In BuilderSolve the schedule is built from the estimate's line items, so a rough estimate produces a Gantt with real phase durations you can set against the jobs already on the calendar. That shows you where the overlap is. It does not tell you who your superintendent will be — that is a name only you can write in. If takeoffs are the bottleneck, the BuilderSolve estimating team can run them, which leaves your time for the decision rather than the counting. What the back-office team cannot do is supervise a jobsite. If the staffing test fails on field supervision, outsourcing the office does not fix it, and we would rather say that here than have it come up after you sign.
What to do this week
Pull every bid invitation currently open on your desk. Run both tests on each one — the calendar overlay with names, and the scope read against the list above. If one fails, call the owner or GC today, decline and give the reason in one sentence. Then put the review on the calendar for the next invitation before the takeoff starts, not after.
Common questions
how do you decide whether to bid a construction job
Run two tests before the takeoff starts. First, can you name the superintendent and crew for every month the job will actually run, including a realistic start-date slip? Second, do the documents define scope well enough to price without guessing at major systems? If either fails, decline or qualify the bid. Whether you can win the job is a separate question and should not drive the decision.
is it bad to decline a bid invitation from a GC or owner
No, not if you decline early and give a reason. A short call or email saying the schedule conflicts with committed work, or that MEP scope is undefined, tells the owner or GC you read the documents. That keeps you on the list more reliably than a late, understaffed job does. Declining on bid day with no explanation is what damages the relationship.
what should a bid no bid checklist include
A useful bid/no-bid checklist asks two things before estimating starts: whether you can staff the job for the months it actually runs and whether the scope is defined enough to price. Most checklists score win factors like owner relationship and bidder count instead. Those predict whether you get the job, not whether it makes money. Keep the win factors secondary to the two tests.
can you bid a construction job with incomplete drawings
You can, but what you send should match what the documents support. Price the defined scope and carry undefined parts as stated allowances, unit prices or written exclusions. A single lump sum over undefined scope moves the design risk onto you, and padding it with contingency usually gets cut in negotiation. If the gaps are too large to qualify around, decline the bid.
what happens if you win a job you don't have the people to run
The job runs longer than priced because a split or newly hired superintendent moves slower. General conditions are a weekly cost, so every added week comes out of margin. The jobs you pulled people from slip too, and their punch lists and retainage drag. The loss is spread across several jobs rather than coded to the one that caused it, which is why it is easy to miss.