theTakeoff.ai/your-missing-millions · prepared with the figures entered on the page
RFP Pros × ContraVault · Capacity benchmark
Your missing millionsWhat no-bids cost you in a year
Most industrial contractors track the bids they lose. Almost none track the bids they never ran.
This counts one thing only: work that passed your filter, that you wanted and were qualified for, and that never got priced because no estimator was free before the deadline.
Fig. 1 · Bid coverage bands · RFP Pros Survey 2026, n = 312
Price your own coverage gap
The capacity cost calculator
Seven inputs. Nothing is sent anywhere: the math runs in your browser, and your results download as a professionally formatted PDF. Defaults are the median US industrial estimating team from the 2026 benchmark.
Your numbers
Work you wanted and were qualified for, dropped because the desk was full.
Median in the 2026 benchmark is 22%. Top quartile is 38%.
CFMA puts specialty contractors just above 16%.
Takeoff plus pricing. Median is 32 hours, top quartile 19. This converts the gap into hours you can staff or automate.
Make this stricter
Set to 75% because work you pass on skews toward lower-fit opportunities. Lower it and every figure on the right falls with it.
Your capacity ledger · annual
On 6 to 7 jobs a year you were positioned to win and never put a number against.
Reaching top-quartile coverage of 68% would mean pricing 16 more packages a year: about 500 estimator-hours, or 12 estimator-weeks you do not currently have. Worth roughly $2.2M in revenue and $350K in gross profit.
First move, not the whole gap: five points of coverage is 3.4 more packages, worth about $75K in gross profit.
Nothing you type is transmitted or stored.
Start measuring it this week
Most estimating desks cannot produce the second input on this page, because nobody logs it. The fix takes two fields on whatever you already use to track opportunities.
Why did we decline this opportunity?
Would we have bid it with two more estimator-days?
The second field is the whole measurement. The next opportunity you decline, log both. Within a quarter you will have a real coverage number instead of an impression, and a figure you can take into a budget conversation.
What a capacity no-bid is
A qualified opportunity declined purely because no estimator was free to price it before the deadline.
A strategic no-bid declines work on merit: wrong scope, wrong owner, wrong risk, wrong margin. That is a decision. A capacity no-bid is a symptom of a constraint, and the two get filed under the same heading in most CRMs.
The distinction matters because the industry only teaches the second one. Search for "bid/no-bid" and you will find scorecards, weighted matrices and go/no-go frameworks, all built to help you decline more work, more deliberately. That advice is sound. A disciplined filter is worth 5 to 10 points of win rate on its own.
But it assumes the constraint is judgment. For most industrial estimating teams in 2026 the constraint is hours. You are choosing between the package due Thursday and the package due Friday, and the one you drop is rarely the one that scored lowest. It is the one that arrived second. The filter said yes. The calendar said no.
The tell. If you can name a job from the last year that you wanted, were qualified for, had a real shot at, and passed on anyway, and the reason was the calendar rather than the scope, that is a capacity no-bid. In the 2026 benchmark, 63% of teams could name at least one.
How the math works
Four steps, no black box. Every figure is either yours or a published benchmark cited at the bottom of this page.
Count the work you never priced
Add the opportunities you declined for capacity to the bids you did submit. That is the pipeline you actually saw, and your bid coverage is the share of it you were able to price.
coverage = bids submitted ÷ (bids submitted + capacity no-bids)
Apply your win rate, then discount it
Declined work would not convert at your normal rate, because opportunities you pass on skew toward lower-fit, later-arriving, more crowded jobs. The model applies a 25% penalty by default, so a 22% win rate becomes 16.5% on declined work. Tighten it in "make this stricter".
expected wins = no-bids × win rate × 0.75
Convert wins to revenue, then to gross profit
Revenue is the headline. Margin is the number your CFO will ask for. CFMA's benchmarker puts specialty contractors just above 16% gross margin, with best-in-class near 21.8%.
lost gross profit = expected wins × package value × gross margin
Convert the gap into hours
Dollars make the case to a CFO. Hours are what a chief estimator can act on. The packages between your coverage and 68% are multiplied by your hours per package, then divided into 40-hour estimator-weeks: the capacity you would need to find, hire, or take off the desk.
hours to close = packages to reach 68% × hours per package
Where this model is deliberately conservative
It ignores follow-on and repeat work from a won job. It ignores the relationship value of showing up to bid at all. It applies a 25% penalty to the win rate on declined work. And it excludes the cost of the bids you did run badly under time pressure, covered in part 05 but never counted in the headline number. The figure the calculator returns is a floor.
The benchmark behind the defaults
If you do not track a number the calculator asks for, use these lines. Once you do have your own, measure against them.
| The metric |
Leading top 25% |
Median | Lagging bottom 25% |
|---|---|---|---|
| Bid coverage, share of qualified pipeline priced | 68% | 45% | 31% |
| Bid win rate | 38% | 22% | 14% |
| Bids submitted per estimator, per year | 41 | 27 | 18 |
| Hours per mid-size package, takeoff plus price | 19 | 32 | 46 |
| Estimate-to-actual variance | ±6% | ±10% | ±16% |
| Rework as a share of estimating time | 14% | 24% | 34% |
| Months to ramp a new estimator to solo bids | 13 | 20 | 26 |
Fig. 2 · The estimating benchmark scorecard · RFP Pros Survey 2026 · hours, variance, rework and ramp: lower is better
Hours per package here is the all-trade median. Process piping and mechanical is the heaviest desk in the study at 38 hours for a mid-size package, which is why it is usually the first place capacity runs out. The full dataset is in The State of Industrial Estimation 2026.
Why the real number is bigger than the calculator says
The headline figure only counts work you never priced. It does not count what the same constraint does to the bids you did submit.
The rework tax
Industrial estimators report spending about a quarter of their time re-checking numbers, chasing missing information, and redoing takeoffs when a drawing or an assumption changes. On a four-person desk that is close to one full estimator's year, spent correcting avoidable error rather than pricing new work. It is capacity you already pay for and do not get.
The accuracy tax
Two in five industrial teams report a typical estimate-to-actual variance above 10%, wide enough on thin-margin work to erase the profit on a job you won. Time pressure is what widens it. The bid you rushed on Thursday is the margin you lose in month four.
is the fully loaded cost of running a single large industrial bid cycle at roughly 875 estimator-hours, at a burdened rate near $110 an hour. That is what the desk spends to produce one number, which is why the hours that go into it and the packages that never get any hours at all are the same problem.
Derived: Constructing Excellence bid-cost survey, BLS cost-estimator wage data, theTakeoff engagement dataNone of this eases on its own. 35% of industrial estimators are 55 or older, the median ramp to solo bidding on a complex package is 20 months, and the BLS projects cost-estimator employment to decline 4% through 2034 while still opening roughly 16,900 roles a year, every one a replacement rather than a new seat. That is a separate argument, made properly in the benchmark report.
What top-quartile desks do differently
Rank the 312 teams by estimating capacity and the top quartile separates on every axis at once, against the rest of the field.
Fig. 3 · Top-quartile firms against the field · RFP Pros Survey 2026 · leaders are also 44% likely to run AI-assisted takeoff against 9% of the field
They industrialize the count
Repetitive quantity takeoff comes off the senior estimator's desk, so judgment time goes to scope, risk and price rather than to counting.
They reconcile across trades systematically
71% of industrial packages now span three or more trades, and multi-discipline bids are won or lost at the interfaces rather than inside any one trade.
They qualify harder, then cover more
A disciplined filter and high coverage sit together. The filter decides which work. Capacity decides how much of it you can actually price.
They measure the desk
Coverage, hours per package and rework are tracked as numbers that move, rather than treated as facts of the trade. That is what makes the two fields in part 01 worth adding.
Where theTakeoff fits
The machine takes the counting. The estimator keeps scope, productivity assumptions, price and final accountability.
theTakeoff is the industrial estimation module inside the ContraVault platform. It reads your drawings, P&IDs for line lists, valves, instruments and spec breaks, isometrics and scaled GAs for lengths, and returns priced quantities seeded on standards including ASME B31.3, B16.5 and MCAA labor units.
Every quantity carries a confidence score, lands in a review queue, and links back to the exact spot on the drawing it came from. An estimator can accept a number or overrule it on the evidence in front of them, in seconds, which is the part that makes a first-pass count usable at all. On a large industrial bid cycle that has meant roughly 875 estimator-hours down to about 300.
Compare that against the hours figure the calculator gave you. The question is not whether a tool can count faster. It is whether the hours it gives back are enough to cover the packages you are currently declining.
Questions estimators ask about this number
What is a good bid coverage rate?
In the 2026 benchmark the median US industrial estimating team priced 45% of its qualified pipeline and the top quartile priced 68%. Below roughly 40%, capacity rather than strategy is usually setting the number: the filter has stopped choosing and the calendar has started.
Why discount the win rate on declined work?
Because assuming declined work converts at your full rate would overstate the number, and any chief estimator would spot it immediately. Work you pass on skews toward jobs that arrived late, sit further from your strike zone, or are more crowded. The model applies a 25% penalty by default and lets you make it stricter.
How much does it cost to prepare a bid?
Constructing Excellence puts the cost of bidding at roughly 0.48% of project value on losing bids and 0.65% on winning ones, and found that firms invested about 25% more in the bids they went on to win. On large industrial packages a single bid cycle runs near $96K at about 875 estimator-hours. The larger cost for most teams is the bid they never started.
How does the 22% median win rate compare with the wider market?
It sits inside the band the rest of the industry reports. FMI research has long put competitive general contractors at roughly one win in five bids, and hard competitive work typically runs 10% to 25%, with negotiated and selective work higher. Survey figures on this page are never blended with third-party figures inside a single number.
Next step
Bring your two fields, we will bring the benchmark
Thirty minutes. Watch theTakeoff run a real industrial package end to end, then walk your own capacity numbers with us: coverage, hours per package, and what closing the gap is worth on your pipeline. If you have started logging declines, bring the log.
Sources and methodology
Practitioner benchmarks are drawn from the RFP Pros State of Industrial Estimation Survey 2026, fielded January to February 2026 to 312 estimating professionals at US industrial contractors across process piping and mechanical, structural steel, electrical and instrumentation, insulation, equipment, and multi-discipline EPC firms. Third-party figures are cited individually below and are never blended with survey data inside a single number.
- US Bureau of Labor Statistics, Occupational Outlook Handbook: Cost Estimators. Median annual wage $77,070 (May 2024), employment projected to decline 4% 2024 to 2034, roughly 16,900 annual openings, all from replacement need.bls.gov
- CFMA, Construction Financial Benchmarker (2024). Specialty contractor gross margin just above 16%, best-in-class 21.8%.cfma.org
- Constructing Excellence, Bid Cost Survey. Cost of bidding at 0.48% of project value on losing bids, 0.65% on winning bids, winners invested about 25% more per bid.constructingexcellence.org.uk
- FMI research, as reported in industry analyses. Competitive general contractors win roughly 1 in 5 bids.bidicontracting.com
- ConstructConnect, Bid or No-Bid. Bid cost components including the opportunity cost of skipped bids, complex commercial bids at 20 to 150 estimator-hours.constructconnect.com
- NCCER and AGC of America. Roughly 41% of the construction workforce projected to retire by 2031, about one in five workers now over 55.agc.org
- theTakeoff engagement data, 2026. Estimator-hours per large industrial bid cycle, before and after AI-assisted takeoff. Internal figure, not survey data.thetakeoff.ai
Model assumptions. Expected wins on declined work = capacity no-bids × win rate × 0.75. Lost revenue = expected wins × average package value. Lost gross profit = lost revenue × gross margin. Weekly figure = annual lost revenue ÷ 52. Packages to reach top-quartile coverage = (opportunities seen × 0.68) − bids submitted. Hours to close = packages × hours per package, at 40 hours to an estimator-week.
Dollar outputs are rounded to two significant figures and win counts shown as a range, because the inputs are estimates and false precision would imply otherwise. Excluded from every figure: follow-on work, relationship value, and the cost of degraded accuracy on submitted bids.