At the beginning of this series, we organized project information around the work. We turned sequences into work packages, defined planned picks, built durations from crew-days and used the schedule to understand resource demand. Then we established rules of credit and compared progress with the hours being spent.
The last step is making sure that experience reaches the next estimate.
A project can finish with a final cost report and still leave the estimating team without much useful information about how the work actually went. We may know the total tons, total hours and final MH/ton. But unless we preserve the conditions behind those numbers, we leave the next estimator to reconstruct the operation or make assumptions.
Keep the explanation with the number.
Suppose a package finishes above its estimated MH/ton. Before carrying that rate into another bid, we need to understand what changed.
Did the package contain more light, fragmented work than expected? Did preassembly change the number of crane picks? Were the crew and crane waiting on access, deliveries or another operation? Did follow-up work require more hours than the estimate allowed?
The relationship from earlier in the series still helps:
Picks/ton describes the work mix. MH/pick describes the labor consumed per pick. Both need context. A repetitive beam pick and a tandem truss pick are different operations, even though each counts as one planned pick.
The labor scope also has to match. Raising-gang MH/pick cannot explain total-project MH/ton by itself. Plumb and bolt, welding, decking, detail, supervision and other support hours need to remain identifiable.
A useful record preserves enough information to explain the result: the work package, quantities, planned and completed picks, crew definition, crane arrangement, duration, labor hours and conditions that affected production.
We do not need to document everything. We need to retain what would change the next planning decision.
Compare the assumption with what happened.
The original plan gives us a starting point. We expected a defined crew to complete a particular work class at an assumed rate under certain conditions.
At the end of the package, we can compare that assumption with actual performance and ask what belongs in the next estimate.
Some differences should change the base production rate. Others belong in a separate allowance or a condition attached to the rate. Restricted access on one project does not mean every future package requires the same additional time. But if the next project has similar restrictions, the estimate needs to recognize them.
I learned this when my initial duration calculations came in shorter than a superintendent’s schedule. After we talked through the work, I understood that I had missed the effect of restricted access. I revised the model to reflect it.
The useful lesson was the relationship between the operating conditions and the time required. Keeping that explanation with the rate makes it usable later.
Carry the process forward too.
The microchip fab example in the rules-of-credit note showed another kind of feedback.
We connected model quantities, the project manager’s schedule and agreed completion gates into a weekly update. The team used that information to review progress, manpower needs and inventory available to feed two cranes. We carried the same process into the second phase.
We did not have all the production tracking shown in the production model examples. We did have a repeatable way to organize the work, update its status and discuss what was needed next.
That is worth preserving alongside production rates. The work-package structure, claiming rules and update routine can all make the next project easier to manage.
The people using the system need to be part of deciding what gets carried forward. Estimating needs the assumptions and results. The superintendent and project team provide the explanation of how the work was actually performed.
Use comparable work to build the next plan.
Over time, those records can become a useful set of references for estimating.
The question becomes: where have we performed similar work, with a similar crew and crane arrangement, under similar conditions?
That gives the estimator a basis for selecting production assumptions and making the differences explicit. Model quantities and planned picks can then be converted into crew-days, man-hours and a preliminary schedule. The team can review that operating plan against the estimate before committing to it.
This also gives us a reason to review completed packages while the job is still moving. The same lesson may apply to a later sequence, another phase or a bid already being prepared. Waiting until final closeout may mean the next decision has already been made.
Put the jobs on the same timeline.
Once individual projects have time-phased resource plans, another question becomes possible:
What happens when we need to build several of them at once?
A backlog total tells us how much work is sold. It does not tell us when that work will need ironworkers, cranes, welders, welding machines or preassembly space.
Two jobs may each have a workable plan and still create a conflict when their resource demands overlap. A delayed start on one project can move its peak directly into another. A prospective bid can look attractive on its own while requiring resources already committed elsewhere.
Connecting project schedules gives the organization a way to see those overlaps and discuss its options. It can examine start dates, staffing, equipment commitments and the work it can realistically pursue.
Awarded work and potential work need to remain distinguishable. A bid is a scenario until it becomes a commitment.
Close the loop.
The purpose of this series is to make the connection from estimating to execution and back more usable.
Organize information around the work. Make the production assumptions visible. Build a schedule the team can question. Track completed scope and labor consistently. Preserve the explanation of what happened so it can improve the next plan.
The tools will vary. The connection is what needs to survive from one project to the next.
The next series will explore that organizational view: combining project demand, backlog and potential work to understand the people, equipment and supporting capacity required across the business.