The Estimating-to-Field Handoff: Where Margins Go to Die

Ask any controller running a WIP review: the most common pattern in subcontracting is the job bid at 8% that closes at 3%, with nobody able to name a single catastrophe that explains it. No disaster, no dispute — just a hundred small overruns: crews sequenced differently than the estimate assumed, premium-time that was never in the labor number, change work performed free because nobody in the field knew it was out of scope.

The industry calls it fade and shrugs, but fade this consistent isn't weather — it's information loss. The estimator spent fifty hours building a detailed model of how this job wins: production rates, crew mixes, sequence assumptions, supplier quotes, exclusions. Then the job was won, the folder changed hands, and the field started building from the drawings and their habits — while the model that justified the price stayed in estimating's head. The margin didn't die in the field. It died in transit.

Walk a fading job and interview both ends; the same gaps appear every time. The estimate priced Area B at a production rate that assumed work off the deck before walls closed in — the field arrived after drywall and hand-carried everything, 40% slower, and nobody flagged it because nobody knew the assumption existed. The estimate excluded premium time — the field worked Saturdays to hold a schedule the GC compressed, unbilled, because the foreman didn't know Saturdays were extra. The bid carried $30K for equipment the field never rented because they didn't know it was in the number — false savings that showed up as labor overrun instead.

And the biggest one: scope. The scope letter and its exclusions — the most-negotiated pages of the whole deal — routinely never reach the trailer. A foreman who doesn't know that fire-caulking at rated penetrations was excluded will do it for free all job long, politely, as 'part of the work.' Every exclusion your estimator fought for is a change order the field can only capture if they know it exists. That's not a legal document. That's a field tool filed in the wrong drawer.

The fix is a meeting — mandatory, before mobilization, estimator plus PM plus superintendent plus foreman, ninety minutes for a mid-size job. Not a folder transfer; a testimony. Agenda block one, the money: the estimate at line-item level — labor hours by area and system, production rates assumed, crew mix, equipment carried, major supplier quotes. The field walks out knowing the number they're being asked to beat, by cost code. A budget the field has never seen is not a budget; it's a postmortem waiting to be written.

Block two, the assumptions and sequence: how did the estimator imagine this job going? Access, laydown, hoisting, predecessor state, phasing. Every assumption is a tripwire: 'if it's not like this when you get there, that's a flag-it-today change event.' Block three, scope boundaries: the scope letter and exclusions read aloud, translated into field language, posted in the trailer. Block four, the risk map: the contract's sharp edges the review flagged — the LD exposure and schedule float, the notice deadlines, the payment terms feeding the cash forecast — each translated into a field behavior: what to watch, when to yell, who to tell.

Two ground rules make the meeting land. First, the estimator presents — not the PM reading the estimator’s file. The person who built the number answers the field’s questions live: why is this rate so hot, what did you assume about hoisting, is there money in here for the second mobilization? Half the value is in the cross-examination. Second, disagreements get resolved in the room, not deferred. If the superintendent says the Area-B rate is fantasy, that conversation happens now — while there is still time to re-sequence, buy equipment, or brief leadership that the job carries a known pressure point — not in month four as an overrun everyone privately saw coming.

Meetings evaporate; artifacts persist. Three one-pagers should leave the handoff laminated. The budget card: hours by major cost code, the production rates that justify them, and the week-by-week manpower curve — the foreman's scoreboard. The scope card: ten bullets of what's ours, ten of what's expressly not, in trade language — the change-order tripwire list. The risk card: this job's specific dangers — 'schedule has zero float in March; any predecessor delay gets flagged same day,' '48-hour notice clause; when in doubt, write it' — the field translation of everything legal review found.

Notice what these artifacts do to your other systems: the daily report gets sharper because the field knows which deviations matter; the change order machine gets fed because recognition was trained on THIS job's boundaries, not generic ones; and weekly cost review becomes a conversation about variances from a shared plan instead of a surprise recital. The handoff isn't another meeting on the pile. It's the ignition for every field discipline you've been trying to install.

The handoff's twin runs in reverse at closeout: the debrief, where actuals confront assumptions. Which production rates held? Where did the estimate's sequence diverge from reality, and what did it cost? Which supplier quotes were fantasy by mobilization? This is the cheapest estimating calibration data on earth — your own jobs, your own crews — and most shops throw it away by never holding the meeting. The estimator who learns their Area-B rate was optimistic by 30% on this building type prices the next one right; the one who never hears it fades the next three bids identically.

Culture warning: the loop only works if it's calibration, not blame. An estimator flogged for every miss starts padding invisibly, and now your bids lose competitively for reasons nobody can see. The tone leadership sets — 'we're tuning the machine, not scoring the people' — decides whether the feedback flows. The shops that get this right develop something rare: estimates the field trusts and fields the estimators trust, which is another way of saying a company whose numbers mean something at every level.

Margin fade is mostly information loss, and information loss has a cure: a real handoff — numbers, assumptions, scope boundaries, risk map — delivered by the estimator to the field in a meeting, survived by three laminated one-pagers, and closed out by a debrief that calibrates the next bid. None of it costs software. All of it costs the discipline to hold the meeting when everyone's busy and the job 'seems straightforward.'

Run the experiment on your next award: ninety minutes, four people, three cards. Then watch what the foreman does the first time a predecessor trade isn't finished in an area the estimate assumed clear — because now they know it matters, know it's money, and know who to tell. That reflex, multiplied across a job, is where the 8% you bid becomes the 8% you keep.