---
id: 2026-05-25T15:02:03-0400
title: 2026-05-25 15:02:03
tags: []
daily: "[[2026-05-25]]"
---
# 2026-05-25 15:02:03

The perceived value of detailed takeoff is baffling
because its single advantage over reference class forecasting
(which offers innumerable advantages over detailed takeoff)
is in itemized pricing,
but this advantage is nullified
by the known phenomenon that the precise
selection of particular items and their quantities 
can not be trusted to be exactly accurate
due to uncertainty of project conditions,
and by the observation that material pricing at the time of bid
is not representative of the cost of buyout.

The inevitable retort is
"It's an *est*imate, not an *exact*imate,"
or some other such reductive dismissal.
True enough, this method is _a_ way
of getting close to the true answer of project cost and schedule,
but it's not an intuitive one.

For an organizational process to be economical
it must contribute value greater than its cost.
Reduction of loss (risk mitigation) is a form of value,
and is the vector by which detailed takeoff is expected to be economical:
a more confident estimate (a narrower high-confidence interval)
allows the organization to bid more aggressively (bid lower)
without increasing their assumed risk in project execution.
A lower bid increases the probability of award,
decreasing the risk that estimating resources will have been wasted.
We can assume that this is not the vector by which _[[conest|ConEst]]_
is expected to generate value,
both because it does not appear to be effective in this regard[^1]
and because the ConEst estimate is not used as basis to decrease the bid.[^2]

[^1]: See [[2026-05-22_11-57-18]].

[^2]: This may be missed opportunity.
    My baseless (and perhaps naive) speculation
    is that a "Hey, we took a closer look and got our price down a bit more"
    may indicate to the customer a [[stag-hunt|stag hunting]] behavior
    that is more valuable than the real decrease in expected profit.

As an estimating method,
detailed takeoff is not more _precise_
(its use does not significantly decrease our certainty),
but it is more _granular_.
An estimate based on detailed takeoff can be easily[^3] subdivided
for use in [[earned-value-management]],
but its most immediate use is in identifying variation
from the expected case.
In other words,
deciding when overruns can be blamed on mismanagement,
and when they are simply bad luck.

[^3]: Depending on the takeoff software.

The last case is where I'd bet the perceived value of ConEst is found.[^4]
I can't help but feel that this effort is misplaced.
If [[pdi-operations|ops]] is like [[pdi-estimating|estimating]],
I have no doubt that they could _all_ benefit from training.
Identifying who exactly needs it more, then, is pointless.
As well to identify which cases of great project success
can actually be attributed to exceptional management
and not simply [[the-failure-of-risk-management#Red Baron Effect|the Red Baron effect]].

[^4]: Crew loading and schedule are important points in estimating,
    and EVM is a fine solution there,
    but it would be straightforward to apply the same RCF principles
    that we use for Bid estimates for these as well.
    That we use detailed takeoff in lieu of such methods
    suggests that it is believed necessary for other reasons.

***

Working exclusively in multifamily for this time
has lead me to spend a lot of time thinking about the question
of how estimating resources are best allocated.
My understanding of the industry is that,
despite the apparent focus on cost,
executive focus is usually on maintaining a consistent volume of work.
Because as contractors we are thinly-capitalized,[^5]
failing to do so seems to represent a larger risk
than a low-margin contract.
Under these conditions,
estimating has very little to do with the project itself,
so we can make assumptions about it
that might not normally seem appropriate.
It is because of our narrow market
that we get to have these much more interesting discussions.

[^5]: See [[merrow_2022_contract-strategies]].