Something odd happens on every plan and spec job. The engineer issues a drawing set, and then the same counting exercise is performed, independently, by almost everyone who touches it.
Everybody counts
Follow one set of mechanical drawings through a bid and you find the same work repeated:
- The general contractor takes off quantities to price the job and test the budget
- The mechanical contractor takes off the same equipment to build their bid
- The distributor takes off terminal equipment to quote it
- The manufacturer's rep takes off what they need to size and price their line
- Somewhere in there, a builder or developer is doing a version of it for their own numbers
Five or six passes over the same sheets. Same diffusers, same schedule, same tags, counted separately by different people in different offices, most of them arriving at slightly different numbers.
Nobody designed this. It is what happens when the information exists on a drawing but not in a form anyone can pass along.
The cost is not the hours
The obvious cost is labour, and it is real: four to eight hours a set, several times over. But the more expensive part is what the duplication does to the bid itself.
Every count is a chance to differ. When the contractor's number and the distributor's number do not agree, someone spends an afternoon reconciling them, or worse, nobody notices until the order is placed. Meanwhile the whole chain is slow, because each party can only start once the one before them has finished.
And plenty of bids simply do not get quoted. If a set lands and nobody has a free day, the opportunity quietly disappears. That is capacity, not competitiveness.
Up and down the chain
We started with distributors because that is where we could be most useful fastest, and because a distributor's takeoff has a tight, well defined scope: terminal equipment, no pipe, no fittings, no labour.
But the underlying problem is not a distributor problem. It is a chain problem. A takeoff that is trustworthy at one link should not have to be rebuilt at the next. The interesting future is not a faster count for one party, it is a count that moves: the contractor, the distributor and the rep working from the same structured quantities, each applying their own pricing and judgement on top.
The part nobody talks about: the engineer
There is an opportunity further upstream too, and it is the one we find most interesting.
The engineer producing the drawings is the only party who knows exactly what they intended. Yet they get almost no feedback about how readable their set turned out to be. Nobody tells them that their schedule split across three tables cost six people an hour each, or that a tag convention only made sense beside their own legend, or that a revision moved a unit type without updating the plan.
A tool that reads thousands of sets can see that. Used well, it could tell an engineering firm which of their drafting conventions are costing the downstream chain time, before the set is issued rather than after. Not to police anyone's drafting, but because a clearer set is cheaper for everyone who touches it.
What we think happens next
Plan and spec work is won on response time far more than most people admit. The desk that can quote everything beats the desk that has to choose which bids to decline. As that constraint lifts, two things follow.
The first is coverage. Teams stop triaging and start quoting the whole pipeline, which changes what a bid desk is for.
The second is that the takeoff stops being disposable. Once quantities are structured rather than tallied on a pad, they can be compared, reused and handed on, and the chain gets to stop doing the same job five times.
We are not there yet. But that is the horizon worth building toward, and it is a good deal more useful than making one person count faster.
