Skip to content
EARTHVOQ

[ EARTHVOQ Intelligence ]

Your company's own history, in front of the next estimate

Not what contractors generally achieve. What your crews achieved, on your last eleven jobs, with the jobs listed so you can open them. When there are not enough of them to say anything, EARTHVOQ says that instead.
Start freeSee pricing

14 days. No card.

Company productionEX-100 · CY

180 median

260 this estimate

Your median across 5 completed activities is 180 CY/hr — this is 44.4% above it.

5 observations across 5 jobs · 1 excluded, no hours recorded

Not an industry average. Yours.

Published production rates vary by soil, machine, operator, haul distance and weather far too widely to price a job against. The only rates worth comparing an assumption to are the ones your own crews produced, on jobs you can go and look at.
Company productionEX-100 Mass excavation, CY

Median

180 CY/hr

Observed range

100–200

Sample

5 activities

Period

Oct 2025 – Jul 2026

The median rather than the average, deliberately. One job that hit rock and made 100 CY/hr pulls an average below four of the five jobs on the list — the median stays where the work actually is.

Every figure opens

Under each number is “how is this calculated?” — the method in one sentence, the sample size, the period, the filters, what was excluded and why, and links to the jobs by name.
How is this calculated?Company production median

Method — quantity ÷ hours per completed activity, then the median of those figures. Only rows recording both contribute.

Sample — 5 observations across 5 jobs

Excluded — 1 × NO HOURS RECORDED · 1 × JOB NOT COMPLETE

The work behind it — Alpha One · Alpha Two · Alpha Three · Alpha Four · Alpha Five

A median a contractor cannot audit is a median he will not bid against, and he is right not to.

Where budget goes

How often each activity finished over its own budget, counted — not diagnosed.
Cost behaviourCompleted work

Mass excavationover budget on 4 of 5 · median 6.1% over

Haul and disposalover budget on 3 of 5 · median 4.4% over

Disposalover budget on 1 of 5 · median 3.7% under

“Over budget on 4 of 5” is a fact about five jobs. “You underprice excavation” is a claim about you, and the four may simply have been the four that hit rock. EARTHVOQ does not make the second kind of statement.

Three things worth checking before you send it

The estimate review runs two kinds of check, and every item says which kind it is.

Deterministic

Arithmetic over this estimate alone. Two unpriced lines are two unpriced lines whether or not you have any history — so this half works on your first day.

Historical

This estimate against your completed work. An aggressive production assumption, a cost code that usually runs over, a margin under what your jobs actually returned, or a line item you carry on most jobs and not on this one.

A missing cost is only ever named because your own past jobs usually carry it — never because other contractors do.

When there is not enough history

Below three comparable observations EARTHVOQ states no median at all. It shows the observations themselves and says the sample is too small — because a confident figure drawn from two jobs is worse than no figure, and it is the kind of number that ends up in a bid.

On a new account the intelligence page is mostly this: what creates a benchmark, how far along you are, and which of your jobs did not count and why. That is a more useful first screen than a chart with nothing in it.

What the analyst can and cannot do

It explains

EARTHVOQ computes the medians, the variances and the forecast; the assistant puts them in a sentence and points at the jobs behind them.

It does not calculate, and it cannot write

No margin, forecast, production rate or cost variance is produced by a model. And no intelligence code path can change an estimate, a budget, a cost or a forecast — that is enforced by a test, not by a policy.

Questions people ask about AI in estimating

Does the AI decide what to bid?+

No. It never changes an estimate, a rate, a margin or a price. It states what your completed jobs did and links to them; you decide what that means for the job in front of you. A number the software put in your bid is one you cannot defend to a customer.

Where does the data come from?+

Your own finished jobs, and nothing else. No industry averages, no benchmarks pooled from other EARTHVOQ customers, and no external dataset. One company’s history never appears in another company’s answers.

What happens on day one, before I have any history?+

EARTHVOQ tells you it has nothing to compare against, and tells you exactly what creates the comparison: approved field days with hours recorded beside quantities, costs posted against the job, and completed jobs. That is a more useful answer than a confident median drawn from two observations.

Is my data used to train a model?+

No. Your figures are computed by EARTHVOQ and used to answer your questions. Nothing about your company trains any foundation model.

Related EARTHVOQ capabilities

The history starts the first day you record hours beside quantities.

Fourteen days, no card. Intelligence has nothing to say on a brand-new account and tells you so — what it needs from you is on the page from the first visit. See which plan includes it.
Start freeCompare plans

14 days, no card. Plans from $79 a month.