Delivery dates are usually the first thing a customer asks about and the last thing the order record explains. A buyer may say the dock only accepts deliveries on Tuesday mornings, and the order may later show a Thursday ship date with no note about the conversation. When the shipment arrives at the wrong time, the team cannot tell whether the customer changed the request or the supplier missed the promise. This protocol tests whether order records preserve the requested delivery window, the promise that followed, and the reason for any change.
Research question and scope
The research question asks whether a sales support order record shows the customer's requested delivery window, the date the request was received, the promise the team made, and any later change with its reason. The scope is one team, one order system, and one fixed review window, not a comparison across industries or carriers. The test measures record completeness, not whether the delivery was fast. It also does not judge the carrier's performance, because a complete record can document a late delivery as clearly as an on-time one.
Why the delivery window is an evidence problem
A delivery window is a commitment with operational consequences, and the order record is the only place where the whole commitment can live. The warehouse needs a pick date, the carrier needs a booking, and the customer needs an arrival window that matches their receiving hours. When those dates are stored in separate systems without a shared reference, each team works from a different version. The customer then receives a shipment that satisfies one record and violates another. A useful record makes the sequence readable without a search through email.
Define the requested window precisely
Write the definition before coding any order, because teams use the word window loosely. Treat a requested window as the customer's own statement of when they want the goods to arrive, including any day-of-week restriction, time-of-day limit, or blackout period. A request for "as soon as possible" counts only if the order also records the receiving constraint that makes speed important. A carrier's estimated arrival is not a customer request. Record the channel and the date the request arrived, since a verbal request and a portal note carry different weight.
Separate the request from the promise
The customer's requested window and the team's promised window are two different facts, and collapsing them hides the decision. The team may promise a later date because stock is short, or an earlier date because a truck has space. Record both, with the date each was set and the person who set it. If the promise differs from the request, write why. A record that stores only the final ship date cannot show whether the team negotiated a change or simply overwrote the customer's words.
Identify where the window is stored
Map every field and document that carries a delivery date before the test begins. The order header may hold one date, the line may hold another, and the warehouse may work from a third. Check the quote, the order acknowledgment, the packing list, and the carrier booking. Note which field a customer actually sees and which field the floor actually uses. The gap between those two fields is often where the record loses the requested window.
Population and unit of analysis
Start with every order released in the review window that has a customer-specific delivery requirement. The unit is one order-destination pair, so an order shipping to two sites produces two rows. Record the account, order number, line, destination, requested window, promised window, and the current scheduled date. Include orders that were later rescheduled, because a second date may have replaced the first. Exclude orders with no stated constraint and note each exclusion with a reason.
Set the review window
Choose a window that covers a full order cycle, such as one quarter, and freeze the dates before coding. A short window may capture only simple orders and miss long lead-time items where dates change repeatedly. A long window may span a holiday or a carrier contract change that distorts the comparison. If a policy or system change falls inside the window, split the result at that date and label both halves. Write the window, the order system, and the export date in the report so a reader can judge whether the finding still applies.
Build the evidence dictionary
Each row needs the order identifier, the destination, the requested window, the request date, and the request channel. Add the promised window, the promise date, and the role that made the promise. Record the current scheduled date, the last change date, and the person who made the change. Keep a field for the reason code if one exists, and a free-text field for the reason when the code is vague. Note whether the customer received the change in writing, because an internal edit is not a customer agreement.
Test design
Use a census when the quarter holds a manageable number of constrained orders and a documented sample when it does not. Freeze the export before coding so later edits do not change the target. Have one reviewer code every row and a second reviewer check a subsample without seeing the first result. Resolve disagreements against the order record and the customer's message, not against memory. If a requested window is missing, mark it incomplete rather than reconstructing it from the final ship date.
Code the window signal
Classify each record by the strength of its evidence. A written customer message that names the window and the destination is strong. A note that paraphrases the request but names the date is moderate. A final ship date with no request behind it is weak. Record the channel and the date for every signal, because a weak note from last week is more useful than a strong one from an unknown period. Keep the classification simple enough that two reviewers apply it the same way.
Compare the order with the warehouse slot
Pull the warehouse appointment or pick slot and compare it with the promised window. Note any difference and whether a documented change explains it. A slot that falls outside the customer's receiving hours is a confirmed failure even if the order record looks complete. A slot that is earlier than the promise may reflect a staging decision rather than a delivery. Record the explanation and its source instead of guessing. The comparison shows whether the record and the floor are working from the same date.
Compare the promise with the carrier booking
Check the carrier booking against the promised window as well, because a booking can drift after the order leaves the team. Note the pickup date, the transit time, and the estimated arrival. If the carrier's arrival falls outside the window, record whether anyone told the customer. A booking that was never reconciled with the promise is a handoff gap, not a carrier failure. Keep the booking reference so a later reviewer can repeat the check.
Check partial shipments and split deliveries
Partial shipments create multiple windows, and a record that stores one date will misrepresent the rest. For each partial shipment, record which quantity arrives when and whether the customer agreed to the sequence. Check whether the remaining balance carries its own promised date or inherits the original one by default. A split delivery that the customer never approved can look complete in the system while failing at the dock. The backorder queue test covers the related distinction between a supplier delay and a customer hold.
Resolve conflicts about who changed the date
Reviewers will disagree when a date changed more than once or when a new coordinator inherited an old promise. Resolve the conflict by applying the record in force on the date of the change, not the record as it looks at test time. If a date was changed without a named person, mark it as an exception and record who discovered it. Do not treat a silent overwrite as a customer-approved change. The difference matters when the team reviews how promises actually travel through the order lifecycle.
Repair the order template
Most gaps come from a template that stores a final ship date but not the request behind it. Add required fields for the requested window, the request date, the promise date, and the reason for any difference. Make the customer-facing window visible beside the internal scheduled date so a reader sees both. Pair the fields with the last customer contact test so a stale request is not mistaken for a current one. The order reconciliation routine shows where the same evidence check belongs before billing.
Change the review habit
A field will stay empty if the order review does not ask for it. Add one question to the release review: what window did the customer request, and what did we promise? Ask the coordinator to answer before the order leaves the team, not after the carrier is booked. Pair the check with the next action test so a date change always carries an owner. Keep the question short enough to survive a busy week, and count the orders that cannot answer it.
Limitations and next action
The test can show whether the record preserves a window, but it cannot prove the customer's receiving team will honor it. A dock may close early, a site may reject a partial load, and a carrier may rebook without telling anyone. The result also depends on access to the warehouse calendar and the carrier portal, which some teams restrict. Pair the finding with a short review of delivery disputes to see whether window gaps cluster there. Use that evidence to choose the first repair.
Report the result
Report the number of constrained orders, the share with a recorded request, the share with a recorded promise, and the share whose current date matches both. Add one strong example and one weak example with customer details removed. Show the gap by destination, order stage, and coordinator so the reader can see where the handoff breaks down. Compare the result with the backorder queue test when dates moved because of stock. End with one template change, one owner, and one recheck date.
Review cadence
Run the test before a peak shipping season, after a warehouse or carrier change, and at least once every quarter. A monthly spot check helps during a busy close when dates move by phone and never reach the order record. Keep the same definitions so results remain comparable across cycles. End each review with one change, one owner, and one recheck date. Archive the export, the customer messages, and the coding sheet beside the finding so the next reviewer can repeat the work.
Sources
The sources below offer general context on business records, financial management, and small business operations. They do not describe a local delivery result or a benchmark for this team, and they were checked on November 20, 2026. Use them as background only and keep the local observation separate from any external claim. The list is not a substitute for the customer's written delivery requirement or the carrier's booking confirmation.
- U.S. Small Business Administration: Manage your finances: https://www.sba.gov/business-guide/manage-your-business/manage-your-finances
- IRS: Recordkeeping: https://www.irs.gov/businesses/small-businesses-self-employed/recordkeeping
- U.S. Department of Labor: Recordkeeping: https://www.dol.gov/agencies/whd/recordkeeping
- NIST Cybersecurity Framework: https://www.nist.gov/cyberframework
- U.S. Census Bureau: Small Business data: https://www.census.gov/programs-surveys/susb.html