Bid or No-Bid: Deciding Which Traffic Control Jobs to Walk Away From

Most traffic control companies do not really make bid decisions. Work arrives, someone estimates it, and the only question asked is what number to put on it.

That is a decision by default, and it has a cost: estimating time spent on jobs you were never going to win, and won jobs that consume crews at a margin that does not justify them. A short screen applied before the estimate fixes both, and it takes about ten minutes per opportunity.

Why bidding everything is not free

Three costs, all of them real and none of them on an invoice.

Estimating time is your scarcest labour. The person producing takeoffs is usually the person who best understands the work — often the owner or the most experienced operations manager. Every hour spent estimating a job you will not win is an hour not spent on the ones you will, or on the crews already working.

A bad win is worse than a loss. Winning a job at a price that assumed a day operation when the restriction window makes it a night operation does not merely produce a thin margin. It occupies crews and equipment you needed for better work, which is the more expensive half.

Bidding everything destroys your own data. If you bid indiscriminately, your win rate means nothing and you cannot tell which categories you are strong in. That is not an abstract loss — it is the information that would let you stop bidding the ones you lose.

The screen

Six questions, before any estimating. If more than one or two answers are bad, decline and say so early — declining promptly is professionally better received than a late no-bid or a padded number.

1. Do we know when we are allowed to work? Not the duration — the permitted hours. A closure restricted to overnight is a different job with a different cost base. If the restriction is unknown, that is the first thing to find out, because it changes the estimate more than anything in the drawings.

This is checkable in minutes where an agency publishes it. South Carolina's DOT, for instance, publishes hourly interstate lane closure restrictions segment by segment and direction by direction — its December 2024 edition restricts eastbound closures on one four-lane I-20 section in Aiken County from 7 a.m. on weekdays, with Friday running to 10 p.m., which makes any closure there a night operation and prices it accordingly. On federal-aid projects the underlying policy driver is the Work Zone Safety and Mobility Rule at 23 CFR 630 Subpart J, which pushes agencies to assess and manage work zone impacts — restrictions being the most direct expression of it. Restrictions vary by agency and are revised, so check the current document for your segment rather than relying on a prior job.

2. Is the traffic control scope actually defined? On subcontracted work the traffic control scope is frequently described in a sentence. "Provide traffic control as required" transfers unbounded scope to you at a fixed price. Either it gets defined before you bid or your price includes the uncertainty — and if the prime will accept neither, that is your answer. See subcontracting traffic control work.

3. Do we get paid, and when? Customer payment history is the most predictive single input you hold, and most companies have it and never look. A customer who pays at 90 days on a job with heavy device rental is financing themselves with your equipment. Retainage compounds it — see retainage and cash flow.

4. Can we crew it without breaking existing commitments? Against your actual schedule, not against a feeling. This is where a company with one crew lead spare says yes to work needing two, and discovers it in three weeks. Crew utilisation is the number that answers this honestly.

5. Is the geography workable? Drive time is unbilled and it compounds daily. A job ninety minutes from the nearest yard costs three hours a day per crew before anyone touches a cone. Occasionally that is fine; routinely it is a structural loss, and it is the calculation behind whether a second yard is warranted at all.

6. Do we know why we would win? Price, availability, relationship, credentials, or specialised equipment. If the honest answer is "no particular reason," the expected value of the estimating time is low. That is not defeatism — it is the same discipline as recognising when a customer already has an incumbent they are happy with.

The four that reliably disappoint

Patterns worth naming, because they recur across companies.

The unlimited-scope subcontract. Covered above and worth repeating, because it is the most common. A fixed price against undefined traffic control obligations is a bet that the prime's plan will not change. It always changes, and then the argument is about whether it was a change order or something you already agreed to provide.

The far job taken to keep a crew busy. Superficially rational — better than idle. But it usually displaces closer work that arrives a week later, and the drive time makes it thinner than idle-plus-availability would have been. Taking it is defensible; taking it without doing that arithmetic is not.

The prestigious loss-leader. A large public job taken at a marginal price to build past performance for prequalification. Legitimate, and it belongs in the prequalification strategy — but it should be an owner's deliberate investment decision with a known cost, not an estimate that quietly came in low.

Emergency and callout work priced like planned work. Callouts have different economics: mobilisation at odd hours, overtime, standby, uncertain duration. Priced on planned-work rates they lose money reliably. Either you have agreed callout rates in your master service agreement or you are pricing after the fact, which is the weakest position available — see emergency callout dispatch.

Track the outcome, or the screen never improves

The screen is only useful if it learns, which requires recording two things per opportunity: the decision and the reason, and — for jobs bid — whether you won and roughly where you were.

That produces the answers no amount of judgment will: which customers you win with, which job sizes suit you, at what distance you stop being competitive, and which categories you lose consistently regardless of price. Most companies that start recording this discover a category they have been reliably losing for years and had never seen as a category, because each loss looked individual.

It also makes strategic bidding honest. If a bid is priced to win a relationship rather than to make margin, record it as such. Otherwise it contaminates the history and your hit rate becomes uninterpretable — and next year someone points at it as evidence you should price everything that way.

What no-bidding well looks like

Decline early, briefly, and without a story. "We're not able to give this the attention it needs on that schedule — please keep us on the list" is sufficient and preserves the relationship better than a padded number, which customers recognise and remember as unreliability rather than as a decline.

The counterintuitive result is that companies which say no more often get invited more often. A contractor whose price you can trust and whose yes means yes is more useful to a general contractor than one who bids everything and occasionally cannot deliver. Which makes the bid/no-bid screen a reputation instrument as much as a margin one — and the account management benefit compounds over years.