Who Signs Off on Pilot Order, and What They Look for
The sections below work through pilot order signoff as a re-order conversation with a supplier you already use. The pilot order is the first document in a sourcing file that asks somebody to put their name against a decision. A request for quotation asks a supplier for a number. A sample order asks for a piece of product that can be held in the hand. A pilot order asks a person to accept a configuration, a standard of acceptability, a shape of cost, and a risk, in writing.
- Read the technical file — check what the first order assumed about pilot order signoff
- Read the supplier record — check the delivery and support record behind pilot-order
- Read the contract terms — check the price and lead-time drivers for four names again
- Treat this as the pilot order signoff sequence, with pilot order as the reference.
That is why the pilot stage carries a sign-off and the sample stage often does not. The signature turns four decisions made by four people in four conversations into one commitment.
An illustrative scenario, assembled from ordinary sourcing work rather than from any named project. A procurement manager has a supplier whose quotation has been compared twice. The specification was frozen against a sample some weeks earlier. The pilot order is drafted and it stalls — not because anybody objects, but because three people each need something different and nobody wrote down who was expected to give it. The technical lead wants the configuration record attached. Finance wants the payment stages spelled out. The operations manager wants to know what happens to the doors if the pilot fails. The order sits for a fortnight and is eventually signed by the procurement manager, who is the only person who signed anything and who did not hold the authority to accept the configuration.
That is the ordinary failure. It is not a dispute and it is not a supplier problem. It is a missing map of who decides what.
Direct answer. A pilot order is normally signed by three to five people, and each of them reads a different part of the same pack and is looking for a different thing. The technical approver checks that the configuration is frozen and identical to the sample that was approved. The quality approver checks that there is a written acceptance criterion and a named route for a batch that fails. The commercial approver checks that every difference between what was asked and what was offered sits on a list with a name against it. The finance approver checks the shape of the cash: which stages, where the risk sits between production and delivery, and what happens to the deposit if the programme stops. The sponsor signs because a programme needs an owner, not because a sponsor can judge a door fit. This page explains what each of those people looks for and what you should hand them. It is not legal advice, and the wording that binds anybody is a structure to take to a lawyer qualified in the governing law of the arrangement and in the seat of the supplier. Commercial figures — order quantities, production windows, payment stages, warranty shape, delivery terms — are per-programme structures a buyer asks about, and they are resolved at the request for quotation page.
Four names on the same order
A pilot order makes four separable decisions, and each has a natural owner inside a buying organisation. It fixes what the product is. It fixes what counts as acceptable and what happens when a batch is judged unacceptable. It fixes the commercial shape, including the difference between a volume promise and a per-shipment size. And it fixes the cash, meaning when money and goods stop being the supplier's problem.
The number of signatures should follow from those four decisions, not from a habit, and certainly not from whoever happened to be in the room when the template was last edited.
Three things get confused constantly, and separating them is most of the work. The signer is the person whose name appears on the order. The owner is the person who holds the decision and can reopen it. The accountable party is the person who carries the consequence if the pilot goes badly. In a well-run file these are three different people, and the order should show all three where it can.
| Role | The decision they actually make | The artefact they need to read | What stops them |
|---|---|---|---|
| Technical approver | Whether this configuration is the one that was frozen | Freeze record, sample identity, change log | A named objection about drift, or nothing |
| Quality approver | What counts as acceptable and what a failed inspection permits | Acceptance criterion, inspection plan, release right | A standard with no action level and no release right |
| Commercial approver | Whether every difference from the ask is recorded and accepted by a named person | Deviation log, quote and its validity | Unaccepted differences, or figures arriving outside the pack |
| Finance approver | Whether the shape of the cash is acceptable to the business | Payment structure, delivery term, exposure summary | A schedule with no answer on reversibility or deposit |
| Sponsor | Whether the programme should proceed at pilot scale | A one-page programme case with a stop option | Nothing, unless the pack is written so that reading it is work |
A pilot order signed by one person is not a failure of governance in itself. It is a signal that the other four decisions were absorbed by whoever signed, usually by someone competent, busy, and not the right owner for three of them. The absorbing signer then becomes the accountable party for a configuration they did not freeze and a cash profile they did not build — a position nobody wants to be in when the first batch arrives and something on it is not right. The catalogue side of this is the smart door locks.
Two questions settle the map before the order is drafted, and they are worth asking at the start of the programme rather than at the end of the approval round. Which of the four decisions can you commit to alone, without another signature? And for each decision you cannot commit to alone, who is the person, by role rather than by name, that can? Role, because names change, and the file should survive the person who gave it.
The technical approver and the configuration freeze
The technical approver is not being asked whether the product is good. They are being asked a narrower and more useful question: is this the same thing we approved?
That question has four parts, and a good approver works through them in order. Identity: which model, finish, credential configuration, accessories, packaging and documentation set is being ordered, and against which unit that was approved and received on which date. Drift: what has changed since the approval, including component, process, firmware and packaging changes communicated but not yet accepted in writing. Fit: whether the configuration works on the actual doors in the actual frames, including the survey conditions the specification assumed. And evidence: whether the documents a buyer needs at commissioning and at renewal are attached to the order rather than promised for later.
The freeze is the artefact that makes this possible, and it is worth being precise. A configuration record names the approved specification and the date. A sample record names the unit — which one, received when, by whom, with what seals intact. A change record names every deviation from the approved state since that date and who accepted it. Without all three, the phrase "as approved" does no work, because it cannot be checked against anything.
The commonest failure at this gate is a technical approver handed a quotation and a photograph. They cannot know whether the photograph is of the unit that will be built, and they decline to sign because the only alternative is to sign on faith. A dated product page extract for the specific model, attached to the order, removes most of that friction, and the general principle is worth keeping: confirm the ecosystem and interface position on the dated page for that model rather than assuming it from a family brochure.
Before the pilot, a deeper method for judging the sample is in How to Evaluate Smart Lock Samples Before Bulk Ordering, and the supplier-side view of approval is in Smart Lock Sample Approval Process for OEM Buyers. The freeze is what those two processes produce and what the technical approver then reads.
The quality approver and the acceptance criteria
The quality approver signs for a standard, not for a product. Their signature says that the organisation has decided, in advance, how a batch will be judged and what will happen when it is judged badly. That decision is worth more than it looks, because it is the only thing standing between a rejection and an argument at the loading bay.
Four things must exist before that signature is reasonable. An acceptance criterion: what a good unit is, stated as something checkable rather than a description of how it feels in the hand. A sampling plan: how many units are examined, which defect classes exist, what the action level is, and who applies it. A release right: what a buyer may do when the batch fails — rework, re-inspection, rejection, partial acceptance — and what must be written down for any of them to be exercised. And a record holder: who keeps the inspection report, who holds the sealed reference sample, and where the retained record lives after the pilot ships.
The question that exposes an incomplete set is short. If this batch fails inspection, what happens? If the answer requires a conversation, the quality approver is right to hold the signature, because that conversation will otherwise happen in a corridor with nobody able to authorise the outcome.
Where the organisation has no quality function, the signature belongs to whoever owns the doors. They carry a bad batch on live openings in a way a spec reader does not, and they are the only person whose judgement of acceptability is grounded in what the doors must do.
The audit material that supports this gate belongs in the same file: Smart Lock Factory Inspection Checklist and Smart Lock Factory Audit Checklist for International Buyers both produce items an approver can point at rather than opinions. Whether a supplier will accept a written acceptance criterion and a named release right is a per-supplier commitment to confirm in writing, not something to assume from a catalogue.
The commercial approver and the deviation log
The commercial approver is the only person in the sequence whose real input is a comparison. They read the ask against the offer, line by line, and their job is to make sure that no difference has been absorbed silently into the working specification.
Differences fall into four kinds, and the kind determines who may accept it. A technical difference changes the product, and the technical approver owns it. A commercial difference changes what is paid, when, or what happens afterwards, and the commercial approver owns it. A documentary difference changes what can be proved later, and it is usually accepted by whoever will have to produce the proof. And a silence is not a difference at all — it is an unanswered question that has arrived dressed as a line, and it belongs in the clarification log rather than on the deviation list.
| Deviation kind | What has changed | Who may accept it | What must be attached |
|---|---|---|---|
| Technical | A dimension, finish, interface option, accessory or packaging item | The technical approver, in writing | The changed line and the reference it was measured against |
| Commercial | Price basis, quantity band, payment shape, delivery term, warranty scope or remedy | The commercial approver, in writing | The alternative that was considered, if there was one |
| Documentary | A promised report, record or declaration replaced by a general statement | The named owner of that document | The gap, written as a gap |
| Silence | A requested line was not answered | Nobody; it is re-asked | The question, the date, and the answer when it arrives |
Two commercial structures deserve to be named explicitly during this review, because they are frequently confused and both are asked about as structures rather than assumed. The volume promise is not the per-shipment size: one is a commitment across a period, the other is the size of a single shipment, and an order that mixes them up creates an obligation nobody intended. And the price escalator is not a hidden increase — it is a clause that says which inputs move the price and how a change gets announced, which is what a buyer should ask to see rather than a number they should react to.
The commercial approver needs the quote with its validity period visible, the deviation log beside it, and every commercial figure left as a labelled blank rather than a number carried around on a slide. What each figure turns into is settled per programme in a written offer letter, and the blank is what lets the structure of a term be judged before the number exists.
The finance approver and the cash profile
Finance is not looking at the price. Finance is looking at the point at which control changes hands, and who carries the risk between that point and the next.
That is what a payment structure actually does. The familiar mechanisms — bank transfer in advance, a letter of credit, documents against payment, documents against acceptance, an escrow arrangement, an open account — each answer the same question differently: at which moment does the buyer lose the ability to withhold the money, and at which moment does the supplier lose the ability to withhold the goods? Neither is better in the abstract. Each moves risk, and the right question for a pilot is which side of the movement the programme can survive.
The deposit and milestone design is where the same logic becomes concrete. A schedule of stages is an insurance arrangement: each stage is supposed to insure the buyer against a specific failure at a specific point, and a schedule where every stage falls due before the next stage creates value is not protection, it is prepayment without leverage.
| What finance asks | What the answer protects | Where the answer has to be written |
|---|---|---|
| At which points do we lose control of the money and of the goods? | Exposure in the gap between production and delivery | The payment structure in the order |
| What happens to money already transferred if the programme stops? | Recoverable value if a pilot or programme is abandoned | The deposit and milestone schedule |
| Is the schedule reversible, and who may reverse it? | The cost of changing plan after the first stage clears | The amendment and acknowledgement chain |
| Which currency prices the order and which currency funds it? | Budget variance between quotation and settlement | The written offer letter and the budget line |
| When does the supplier's risk in the goods pass to us? | The position if goods are damaged or lost in transit | The delivery term and the insurance allocation |
| Who carries the cost of a failed inspection? | The cost of rework, re-inspection or rejection | The acceptance criterion and the release right |
The last two rows are where finance and quality meet, and a pack that separates the two views creates a gap that only shows up when a batch fails. Finance is also the person who should be asked early whether the programme can absorb a stop at any stage — because an answer of "not really" changes what the pilot is for, and a buyer who learns that at stage one has lost the option to design around it.
The sponsor signature and what it does not carry
The sponsor is the person whose name goes on a programme. They sign for reasons that have nothing to do with door fit, and asking them to read a specification is the most reliable way to turn a signature into a delay.
What a sponsor can judge: whether the pilot is worth doing, what the programme commits to if it succeeds, what it commits to if it fails, and whether the money is better spent here than elsewhere. What a sponsor cannot judge, and should not be asked to judge: whether the lock fits the door, whether the credential model suits the population, whether the sample is representative, or whether the batch standard is adequate.
The pack the sponsor receives should be one page and should answer five questions. What is this pilot for, and what does it prove. What does it not prove, stated explicitly, because the gaps in a pilot are what people argue about later. What the programme commits to if the pilot passes, and what it commits to if the pilot fails, which is usually a small amount and should be easy to say. And by when the decision has to be made, because a sponsor asked to decide whenever is a sponsor who decides whenever the request suits somebody else.
The trap is the sponsor approval given on a summary that nobody read carefully. It is recorded as an approval, cited at renewal as an approval, and in fact was a signature on a document assembled by the person who wanted it signed. A sponsor signature obtained this way does not protect the programme, because the person who bears the outcome never looked at it.
The second role of the sponsor is the fallback. In many files the sponsor ends up signing not only for the programme but for whatever nobody else would own — a tolerance nobody could resolve, a documentation gap, a delivery term the commercial approver could not judge. Each time that happens the map has a hole and the sponsor is filling it. Counting the fallback signatures in the first pilot is a cheap and revealing exercise.
How the request travels from draft to disposition
The order of the reviews should follow the order of the dependencies, and the dependencies are not the ones most files assume. The configuration has to be frozen before the commercial terms can be judged, because a deviation cannot be classified as technical or commercial until the approved baseline is known. The commercial shape has to exist before finance can judge the cash, because the cash profile is a consequence of the terms rather than a parallel discussion. And the cash profile has to exist before the sponsor is asked, because a sponsor asked to approve a programme without knowing its exposure is being asked to approve an idea.
sequenceDiagram
participant Procurement
participant Engineering
participant Quality
participant Finance
participant Sponsor
participant Supplier
Procurement->>Engineering: Pilot configuration with freeze record attached
Engineering-->>Procurement: Technical concurrence or a named objection
Procurement->>Quality: Acceptance criterion and inspection plan attached
Quality-->>Procurement: Concurrence with the release route named
Procurement->>Finance: Deviation summary and payment shape attached
Finance-->>Procurement: Approval or a condition to be cleared
Procurement->>Sponsor: One page programme case with the stop option
Sponsor-->>Procurement: Approval for the pilot scope only
Procurement->>Supplier: Pilot order issued for acknowledgement
Supplier-->>Procurement: Order acknowledgement with any exceptions
The rule the diagram encodes is that each reviewer receives a summary prepared for that reviewer, not the whole pack forwarded to everyone. One forty-page pack read by six people is read by nobody, and the approver who reads nothing signs because the alternative is to stall. Six one-page summaries cost less to produce and get read.
Two habits make the sequence hold. Each approver is given a date, and the date carries a consequence attached in advance: on that day the item moves to the named next level. A date without a consequence is a suggestion, and a suggestion is what most approval rounds already have. And a reviewer raising an objection is asked to write it as one — what is wrong, against which document, and what would resolve it — so the answer can close it rather than restart the discussion.
The three dispositions and the fourth that is not one
An approval round produces one of a small number of outcomes, and each carries a different weight. Most files lose their authority at this point because the outcome is recorded as a single word in an email.
| Disposition | What it commits you to | What it does not commit | What must be written down |
|---|---|---|---|
| Approved as submitted | Placing the pilot order on the terms in the pack | That the pilot will be repeated or scaled | The pack version approved and the date of approval |
| Approved with conditions | Placing the order once each named condition is closed | Anything beyond the scope of the pilot | Each condition with an owner, an evidence and an expiry |
| Returned for revision | Nothing yet; the request re-enters the sequence | That the objection was accepted or rejected | The objection, the evidence behind it, and who answers it |
| Acknowledged without comment | Nothing, and that silence is the risk | Any obligation on either side | Who acknowledged, when, and whether silence was meant as consent |
The fourth row is the one that causes trouble later. An approver who reads a pack, has no objection, and says nothing has not approved anything; they have merely not objected. The practical difference in a pilot order is small, but it is recorded in the file, and it is the difference between a decision and an absence.
A returned request is a normal outcome and should not be treated as a failure by the person who raised it. The discipline that helps is that a request comes back with a named objection and a named condition for clearing it, so the person who raised it can see it closed. Objects returned as a general unease simply return again, and the order ages in the file.
The approvers' questions can be prepared before the pilot rather than reconstructed afterwards; the evidence each of them needs is set out in Internal Approval Pack for Pilot Order, and what to attach to each submission is handled in Evidence to Attach to Pilot Order.
Conditions are not notes
A condition is an instruction that blocks release until it is satisfied. A note is context. The difference is small in wording and enormous in effect, and the whole point of naming conditions separately is that a condition has an owner, an evidence and a date.
Four things make a condition real. What has to happen, stated as an action rather than a desired outcome. Who does it, by name or role. What evidence closes it — a document, a response, a revised line, a test result. And what happens if it does not close: the order does not go out, or it goes out with the gap recorded and accepted above the approver who raised it.
Conditions written as aspirations fail in predictable ways. The condition with no evidence of closure never gets closed because nobody knows what would count. The condition cleared verbally looks closed and is not, and the person who cleared it has moved on by the time the question returns. The condition inherited by the next buyer at renewal happens when it was never attached to anything that survived the pilot.
There is an asymmetry worth weighing. A condition discovered at approval costs a delay and a round trip. The same condition discovered at the first-article stage costs a rejected or reworked batch, a schedule already committed to installers, and an argument about who knew. Conditions are cheap when written and expensive when discovered, which is the opposite of how they usually behave, because a condition nobody can close is easier to write than one somebody has to satisfy.
The practical control is a conditions register: one line per condition, with the wording, the owner, the evidence required, the date it was raised, and its status. It is short enough to attach to the order, and it is what the renewal approver reads to find out whether the pilot was signed off or merely tolerated.
When the usual approver is unavailable
Authority travels with a documented delegation, not with a rule about who happens to be in the building. The pilot is exactly when the usual approver is on leave, because pilot orders tend to be raised once a programme has just started and attention has moved to the next thing.
Five questions have to be answered before the pilot, and they are cheap to answer in advance.
| The question | Why it matters | Where the answer lives |
|---|---|---|
| Who may give the technical concurrence in the approver's absence? | A concurrence from someone without standing is a comment | A named deputy in the delegation register |
| At what threshold does the sign-off move up a level? | Thresholds come from your own governance, not the supplier | The approval matrix attached to the order |
| How long does a delegation last? | Open-ended delegation becomes permanent authority by accident | An end date on the register entry |
| What happens to approvals given by a deputy when the approver returns? | A good part of the pilot may have run without the owner | A line in the next review note |
| Which requests escalate when an approver is silent? | Silence is the most common cause of a stalled pilot | A named next level and a trigger date |
Recording both parties in the signature block solves a recurring ambiguity. A line saying the technical concurrence was given by a deputy on behalf of the engineering manager records two facts: that someone with standing stood behind the decision, and that the person who owned it was not in the room. A signature naming only the deputy loses the first fact. The reverse ambiguity — an approval given by someone with no standing at all — may be a formality or a real problem, and a register naming deputies with dates is what makes the two distinguishable at a glance.
Escalation deserves one more rule. Decide the trigger before you need it — which level of spend or which kind of objection moves a request up — and name the person who receives it. Escalation improvised under pressure tends to produce a fast decision by the person who is nearest rather than by the person with the standing.
The supplier's counter-signature
The pilot order is not the last document in the motion. The supplier acknowledges it, and the acknowledgement has consequences: it may start the production window, it may confirm or qualify the commercial terms, and it will often attach conditions of its own.
Read the acknowledgement for three things. What it confirms, which should match the pack without drift. What it qualifies, usually a note against a delivery term or a documentation item. And what it adds, which is where the exceptions live. Supplier exceptions are the lines most likely to be skimmed, because they arrive after the signature the buyer has been waiting for and they read like housekeeping.
The acknowledgement chain is also the document that matters most on a repeat order. When the same programme comes back months later, the terms in force are not those of the original request or even of the original order — they are the terms of the order as amended and acknowledged, item by item. A buyer who cannot produce the chain cannot demonstrate what was agreed on the day something went wrong. The order document itself is covered in Smart Lock Purchase Order Checklist; the chain is the part that is easy to leave out.
The supplier also signs on the technical side. Drawings, tooling confirmation, the agreed packaging, the documentation deliverables that will accompany the shipment and the reference unit that defines the approved appearance all need a supplier signature at some point, and several will not travel with the purchase order. A pilot order arriving with no supplier drawing approval is not wrong, but the technical approver will want to know where that approval will be, and when.
One rule prevents most of the difficulty: when the supplier sends exceptions, treat them as a second sign-off round with the same discipline as the first. A returned exception list with a named owner and a date closes. One replied to in a hurry becomes the specification.
What changes when you come back for the re-buy
At renewal the roles are identical and the evidence is different, because the renewal approver is not approving a new product. They are approving a continuation, which means the question has changed from "should we do this" to "what did the pilot prove, what did it fail to prove, and what has changed in our own demand since".
That shift is why the pilot sign-off sheet matters more than it appears. The sheet from the pilot is not paperwork to be filed; it is the best evidence available at renewal, because it is the only document that records what was assumed at the time, which conditions were attached, and what the organisation thought it was buying. A file that holds the pilot disposition, the closed conditions register, the deviation log from the pilot and the performance record can be re-argued; a file that holds only the purchase order has to be re-litigated.
| At the pilot sign-off | At the renewal sign-off |
|---|---|
| The configuration freeze and the identity of the approved unit | What the pilot showed about that configuration in service |
| The deviation summary and the open items | Which deviations recurred and which were closed |
| The acceptance criterion and the inspection plan | Performance against the criterion in real use |
| The conditions register, open | The conditions register, closed, with the outcome of each |
| The cash profile shape | The change in the cash profile since the pilot |
| The stop option | The exit path and what it now costs to exercise |
What changes most between the two moments is the sponsor. At the pilot the sponsor approves an experiment; at renewal the sponsor approves a commitment, usually with a larger number attached, and the question they ask is the one that should have been built into the pilot sheet: what have we learned. If the pilot sheet names the assumptions, the answer is in the file. If it does not, the answer has to be assembled from memory by whoever ran the pilot, which is exactly the person who no longer holds all of it.
The records side of the renewal is covered in Records to Keep After Pilot Order, and the scale question that usually runs alongside it is in How Many Smart Locks Should I Order for a Pilot Project?.
The questions each approver should be asked
Each approver should leave the review having answered one question, written down in their own words. A pack that asks six reviewers to form a view of a forty-page document produces six vague positions and one delayed order. A pack that asks each reviewer a single, answerable question produces a file that can be read later.
| Role | The one question they must answer | The document that answers it | What a weak answer sounds like |
|---|---|---|---|
| Technical approver | Is this the configuration we froze, or has anything drifted since? | The freeze record, the change log and the sample identity | "The same as approved", with no unit and no date |
| Quality approver | How will a batch be judged, and what happens if it fails? | The acceptance criterion and the inspection plan | A sampling rule with no action level and no release right |
| Commercial approver | Where do the offer and the ask differ, and who accepted each difference? | The deviation log | A list of differences with no accepter and no evidence |
| Finance approver | At what point do we lose control of the money and of the goods? | The payment structure and the delivery term | A single line reading "per the quotation" |
| Sponsor | What does this pilot commit us to, and what does it leave open? | The one-page programme case | A signature on a summary nobody read |
Make refusal cheap. A reviewer who says no is doing the organisation a service, and the cost of a no should be a short exchange rather than a delay. A reviewer who cannot say no, because saying no triggers an escalation nobody explained, will find another way to express the same objection.
Where the signature runs into law
This page describes what a sign-off decides and who is asked to look at what. It stops short of the legal consequences, and the gap is deliberate.
Whether a person who signed has the internal authority to bind the organisation is a question about your own governance, and it is worth asking before the first pilot rather than during a dispute. Whether a purchase order and a supplier's acknowledgement together form a contract, and what happens when each document contains different terms, is a question for the governing law of the arrangement. So are the ratification of an approval given without standing, the enforceability of a written acceptance criterion, the weight of a retained reference sample as evidence of an agreed finish, and whether an internal right to see production and quality records is worth anything outside the relationship it was written into.
CONFIRM-WITH-LOCAL-LEGAL: who in your organisation may bind the company on a pilot purchase order, whether an approval given by a delegate or by someone without standing is effective, and how the delegation register should be drafted so that it does what you intend.
CONFIRM-WITH-LOCAL-LEGAL: how the purchase order, the supplier's acknowledgement and the order amendments interact under the governing law, and which document should be the one that carries the commercial terms.
CONFIRM-WITH-LOCAL-LEGAL: whether a written acceptance criterion, a sampling plan and a release right are enforceable as written, and what the inspection record must contain for it to be usable later.
CONFIRM-WITH-LOCAL-LEGAL: what a retained and sealed reference sample can and cannot establish about an agreed finish, and how its custody should be recorded so that its evidential value survives the pilot.
The practical effect of routing these is simple. The approval structure — who reads what, in which order, with which conditions attached — is an administrative design that you are well placed to make, and it should be made deliberately. The wording that binds anybody is drafted with counsel who know the governing law and the seat of the supplier, and brought back as final text.
The sign-off sheet you can keep
The output of all of this is one sheet per approver and one master sheet for the file. The master sheet does not repeat the pack; it points at it, records the dispositions, and carries the conditions register.
| Block | What it carries | Why it is there |
|---|---|---|
| Reference | Project, programme, supplier, order reference, pack version, date | Makes every copy the same document |
| Configuration | Model, finish, interface configuration, accessories, packaging, documentation set | Lets the technical approver answer identity alone |
| Deviations | Number still open, and where the log lives | Stops a clean sheet from being read as a clean quotation |
| Acceptance | Criterion, sampling plan, action level, release right, record holder | Gives the quality approver something to approve |
| Cash | Payment stages, delivery term, exposure summary | Gives finance a shape without reprinting the commercial figures |
| Conditions | One line each, with owner, evidence and status | Makes an approval with conditions trackable |
| Disposition | The outcome, the signatures, the dates | Produces the record the renewal will need |
Three habits keep the sheet honest. Every commercial figure stays in one place, the written offer letter, and the sheet carries a labelled blank pointing at it instead of a number — which is what lets the same sheet be reused for the next programme. Every version of the pack carries its version on the sheet, because an approval given against a version nobody can identify is not an approval of anything. And the sheet is assembled by the buyer, because the buyer is the only party with a reason to know what all five approvers actually checked. A file assembled by the people who approve it will contain the evidence each approver wanted to see; a file assembled by the person who runs the motion will contain what all of them needed, and it is the one that survives a renewal two years later.
The figures are not on this page because they are not fixed. Order quantities, production windows, payment stages, warranty shape, delivery terms, the choice between insurance and an open account — each is a structure a buyer asks about, each is settled for one programme against one supplier, and each belongs in a written offer letter rather than on a webpage. The request for quotation page is where that conversation starts, and if the configuration behind this signature has never been in your hand, requesting a sample order is the step that comes before it.

