The work moves faster than the record.
A wall moves after layout. A soffit gets revised. Someone asks for backing that was not in the package your estimator priced. Your crew can act on the direction before the office has enough information to decide how the extra should be documented and priced.
That gap is where a potential change can stall. The field note is in a message. The photo is on another device. The drawing revision is in the project folder. The person who can explain the sequence is already on the next problem. Each piece may exist, but nobody has a complete package.
This is a systems problem. Telling the PM to stay on top of it does not create the missing record or assign the next action.
Separate four questions that get mixed together.
Was the potential extra captured? Has somebody reviewed and priced it? Was the package submitted and answered? Did the approved amount reach billing? Those are separate questions. A single status called open cannot explain which decision is holding up the item.
Give each gate an owner. The foreman supplies the field record. The authorized reviewer checks scope and supporting evidence. The pricing owner prepares the commercial package. The PM controls submission and follow-up. Accounting receives the approved record with the reference it needs.
Your company may combine those responsibilities in fewer people. That is fine. The handoffs still need names, records, and an action that can be checked.
Use your own log to measure the queue.
Start with actual potential changes from active jobs. Group them by the next action required, not just the job. Add the dollar values of items ready to submit. Separately add approved changes awaiting billing. Do not count an unpriced or disputed item as collectible cash merely because it appears in the log.
Record the age of each item from the event that matters: capture, pricing readiness, submission, or approval. Dollars multiplied by days produces dollar-days of exposure. It does not produce a cash balance. If you estimate financing expense, use your actual funding rate, amount, and period, with those assumptions shown.
The useful result is a ranked queue. Which item needs evidence? Which needs a reviewer? Which has been approved but never handed to accounting? Those are actions your people can take.
Build the record at the source.
A field capture should identify the job, date, location, direction, source, and supporting records. It should also say what is missing. Do not let an automated draft quietly invent a quantity, labor rate, signature, or statement of approval.
Connect the source to the potential change so later correspondence has somewhere to go. Keep the current drawing reference available. A reviewer should be able to reconstruct the basis of a submission without relying on a summary as the only evidence.
When a response arrives, connect it to the same item and assign the next step. A response that asks for backup needs a different action from an approval that needs a billing handoff.
Automate preparation. Keep the commercial decision reviewed.
A scoped workflow can organize records, flag missing inputs, prepare a draft, and surface a waiting item. Your authorized people decide entitlement, price, and submission. The workflow should explain what happens when those people are unavailable or when records conflict.
Start with one representative change and follow it from the field to accounting. That is the work of an Operations Review: map the handoffs, identify where to investigate, and rank the first fixes. The audit determines the build. The goal is a record that can move with the work, instead of chasing it at closeout.