Bid Leveling in New Construction
A requirement written once and reused well saves more time than a clever quotation, so bid leveling new construction is given as a working resource. Every new-build tender is compared against something that has not been built yet.
Direct answer. Short version — bid leveling new construction is treated here as bid leveling new construction first. bid-leveling comes second. The evidence for bid leveling new construction is named in the sections below. A bid leveling new construction gives the specifier working documents. bid leveling new construction is the document that decides the requirement. That is what bid leveling new construction turns into.
On a refurbishment the baseline arrives with a survey. Somebody has stood in the building with a tape, the leaves are there, the frames are in the wall, and two quotations can be read against a shared reality. On a new build the baseline is a door schedule, a hardware specification and a set of drawings, and the leaf that will carry the lock is cut several stages after the award is signed.
That difference decides how bid leveling has to be run here. Bid leveling in new construction is not a reconciliation of price structures. It is the work of forcing every bidder to state, in the same words, the scope they have priced against the same predicted door set. Once that basis is common, the comparison is arithmetic. Until it is common, the comparison is a comparison of two different products.
The working rule. On a new build the schedule's silence is the raw material of every difference between bids. Where the document speaks, bidders agree, because they are reading the same sentence. Where the document is quiet, each bidder fills the gap with the assumption that makes their own offer coherent.
Three parties need the leveling, and they are rarely the three people who wrote the tender. The client wants an installation that works and a building it can defend at handover; the installer wants a lock that fits the leaf that arrived and the frame that is in the wall; the supplier wants to ship what was asked and be paid for what was asked. This article is written for the door hardware consultant who owns the exercise, and for the reader who will have to run the same comparison later without that consultant in the room. Everything in it is illustrative: a method, not a record of any real project, client, company or authority.
A drawing is not a survey, and the difference runs
The sequence on a new build is familiar. Drawings reach the level of completeness the tender can use. The tender goes out against a door schedule and a hardware specification, bidders price against them and return offers in whatever format their quotation systems produce, and the award releases a door order, and the leaf is manufactured to a cut list agreed in the same window. The build options sit in the smart door locks.
Three consequences follow, and each one changes the leveling method rather than merely adding a step to it.
The first is that the geometry in the bid is a prediction. The bidder has no door to measure, so every dimension quoted comes from a schedule cell or a drawing annotation. Two bidders reading the same cell differently produce two different doors, and the difference does not surface until the leaf is cut.
The second is that the award commits far more than the lock. A door order release ties the leaf to a cut list, a handing and a hardware schedule, and it ties the frame to a rebate and a door stop set out before the pour. Everything the award implies is therefore priced into the wall before anybody has fitted a lock to it.
The third is that the schedule's silence is where the money moves. A tender line that reads as a general instruction has asked several questions at once and answered none of them, and each bidder fills the gap with the assumption that makes their own offer coherent.
A working definition is worth fixing at the start: leveling turns a stack of incomparable offers into a documented comparison on a basis the client has seen and accepted. The companion treatment of the term for readers who meet it inside a contract is in Smart Lock Glossary: Bid Leveling Definition for B2b Buyers.
Where two new-build offers part company before anyone reads a price
The separation is almost never in the headline number. Five points account for most of the distance between two new-construction bids.
- Scope of supply. Locks delivered to site, or locks installed and trained, or locks installed, programmed and handed over with the client's own system carrying the access data.
- The geometry assumed. Leaf thickness, handing, whether the bore is cut to the door manufacturer's standard list, and whether the faceplate and strike come with the door package.
- The system scope. Controller, network point, cabling, the power arrangement, commissioning and enrolment labour — one line in a small scheme, a subsystem with its own tender in a mixed-use development.
- Software, credentials and documentation. What the client owns at the end, what is licensed per door and per user, what is hosted, and what submittals, drawings, training records and as-built packs are included.
- Service. Who the client calls, how a failed lock is replaced, who holds the consumables, and whether spares are ordered with the building.
The tool that finds this quickly is a silent-cell register. Walk the schedule and the specification line by line and mark every cell that fails to answer one of the points above. For each marked cell, write the question that should have been answered. Those questions become the clarification request, the clarification request becomes the declared basis, and the comparison becomes defensible because it was built out of the tender document rather than out of the offers. The reviewer's task at this stage is not to find a reason to prefer one bid. It is to find the assumption that separates them.
Eight elements a new-build tender has to price on separate lines
A single extended rate per door set fails on a new build because it hides everything below inside one figure. A bid that prices a lock as fitted and commissioned tells you nothing about whether the bore was cut by the door factory, whether the system was programmed, or whether a handover pack exists at all. The tender therefore has to name the elements that are priced separately, and the bidder has to answer each one. This is not price discovery and it is not a commercial demand; it is the mechanism that makes an offer legible.
| Element the tender must name separately | What the bid has to state | The question that separates two bids |
|---|---|---|
| Lock supply | Variant per door set reference, finish reference, quantities | Which lock, on which door reference |
| Door and frame preparation | Who cuts the bore, who adjusts for backset and faceplate, against which cut list | Is the factory drilling to our list |
| Closing edge and frame hardware | Whether faceplate and strike are supplied, and who matches them to the rebate | Who owns the mismatch between package and lock |
| Network and power | Controller, network point, cabling, and what keeps the door operable on a supply failure | Is the network a supply item or a subcontract |
| Software and credentials | What the client owns, what is licensed, how credential records are handed over | Subscription the client carries, or a platform someone else runs |
| Commissioning, programming and training | Who enrols the first credentials, who trains operators, what is tested | A working system handed over, or a delivered box |
| Documentation | Submittals, dimension drawings, installation information, as-built records | Will the record exist at handover in usable form |
| Service and spares | Support route, replacement path, consumables held | Who answers when a door fails on a Friday evening |
Four of these eight carry most of the disputes. Preparation and the cut list moves the largest amount of money for the smallest change in wording: if the tender does not say who drills to whose list, the door factory drills to its own standard, the lock supplier assumes a prepared leaf, and the installer finds the mismatch on site. That mismatch is not a defect in either product; it is an unnamed join. Commissioning and programming is the element most often assumed to be free because it appears in nobody's scope document, but it is the point at which devices become something the client can operate.
The rule for an element a bidder will not answer is short: a refusal to state an element is an answer. It tells you the bidder has priced nothing for it and that whatever the client receives will be found elsewhere. Record that on the sheet rather than resolving it by assumption, because an assumption made by the reviewer is invisible to the client who signs the award. The published guidance on quotation content is useful background for the same reason: see What Should Be Included in a Smart Lock Quotation? and Smart Lock Quotation Comparison.
The door set is the only unit two offers can share
Most new-construction bids are priced against the lock, and most leveling exercises begin from the wrong object. A line reading smart lock, model X cannot be compared with a line reading smart lock, model X, door set reference D-114, including strike, faceplate, cylinder and cylinder installation, because the two cover different numbers of doors and different amounts of work.
The unit of account is the door set: leaf, frame, hardware and locks, together with the egress route and the locking arrangement of that opening. Pushing every bidder onto door set references from the same schedule, then comparing like with like, is the most productive thing a consultant can do.
| Field that must appear on a bid line | What it fixes | Who has to confirm it |
|---|---|---|
| Door set reference and door number range | Which openings the line covers | The architect's team, against the schedule |
| Door type and the door's own designation | Which leaf, handing and thickness assumption | The door manufacturer, against the door order |
| Lock variant and finish reference per door type | What is physically on the leaf | The bidder, confirmed on the dated product page for the SKU |
| Faceplate, strike, cylinder and their installation | Who supplies the closing edge and who fits it | The bidder, against the door package |
| Egress role and locking arrangement | Whether the door sits on an escape route and how it releases | The fire strategy consultant and the signing authority |
| System scope reference | Whether this door is networked, standalone or excluded | The consultant, against the network design |
There is a second reason the door set is the right unit, and it is procedural rather than commercial. A schedule cell is a commitment: when the client signs the award, that schedule is what the door factory will build against and what the installer will work to. Any leveling exercise that changes a schedule cell has to change the schedule itself, in writing, with a date, and notify every bidder who priced against the earlier version. Otherwise the leveling has silently created a tender that nobody answered. This is the coordination problem in its concrete form, and Bid Leveling: Coordination Between Architect and Supplier treats the working relationship that has to carry it.
Geometry is decided before the pour and written into the order
Of the standing checks a specifier runs against a submittal, the first catches most new-build bids: does the dimensional information describe the door that is actually going to be on this project, rather than the door in the catalogue? A submittal with a brochure and no dimensions cannot be checked at all.
The trap is worth naming precisely, because it is the most common defect in a new-build bid. A supplier quoting the dimensional drawing of a catalogue variant has described a door that exists in a factory, not a door the client has ordered. The dimension the installer needs is the dimension of the door in the door schedule, stated in the format the door manufacturer uses — the same format that goes on the door order. Where a supplier's dimensional data is the only source available, it is a proposal about fit rather than evidence of fit, and it has to be re-checked against the ordered leaf.
| Dimension | What it decides | The format the bidder must use |
|---|---|---|
| Backset | The distance from the door edge to the centre of the spindle or cylinder, and whether the lock body sits correctly in the leaf | A dimension on the drawing for the exact variant |
| Bore diameter | Whether the case passes through the leaf without weakening it or disturbing fire-rated assembly evidence | A dimension, with the tool or template named |
| Bolt-to-bore distance | Whether the locking components land on a stile that can carry them | A dimension relating the bore to the edge and the thickness |
| Faceplate and strike | Whether the closing edge halves match and the strike sits correctly in the frame | A drawing and an item list, never a photograph |
| Leaf thickness and edge detail | Whether the case and faceplate suit the ordered leaf | The door manufacturer's leaf specification |
| Handing and swing | Which variant is offered for which door, and whether the handing was priced correctly | A variant-per-door-type list |
On a new build, half this geometry is settled long before any hardware is specified. The frame is set out in the wall before the pour, and that set-out fixes the rebate depth, the door stop and the position of the closing edge relative to the finished wall face. The leaf thickness is chosen by the door manufacturer and can still move: an acoustic requirement, a fire requirement or an availability change can alter it after the tender and before the order. Handing is fixed when the door order is released: one of the cheapest items on the schedule to change right up to that moment, and one of the most expensive immediately after it.
A dimensional cell in a bid should never carry a number on its own. It should carry the variant the number belongs to and the date the number was true. Model-level facts are never asserted from memory or carried over from a previous tender: they are confirmed on the dated product page for the specific SKU, every time the specification is issued and every time a bid is levelled. Two published references handle this from the two ends: Smart Lock Backset Measurement covers the dimension installers most often get wrong on site, and Smart Lock CAD Drawing Requirements covers what a dimensioned drawing has to carry before anyone can check it against a door.
Three joins where two supply chains meet without a contract
Scope boundaries get crossed in new construction at the joins, and three matter for door hardware. Each has the same shape: two supply chains meet, one assumed the other would act, and neither contract names the join.
The join between the door and the lock. The door factory cuts to a cut list; the lock supplier has a cut list of its own, and the two are not the same document. Where the tender is silent, one bidder has assumed the leaf arrives ready to receive its lock and the other has assumed the leaf arrives raw, and the gap between those readings is worth more than most of the differences the review will later focus on. The leveling instruction is narrow: ask each bidder to name the cut list it priced against, the party it assumes performs the bore, and what happens to the door if that assumption is wrong. The answer tells you which bidder is holding the risk of a cut in the wrong place.
The join between the lock and the frame. The frame, its strike boxes and its door stops usually belong to the door package; the lock, its faceplate and its cylinder belong to the hardware package. The rebate in the frame, the profile of the strike and the thickness of the door stop are properties of the door, and the closing hardware has to live with them. A faceplate that does not sit flat on a rebate produces a door that closes badly and a client who blames the lock. Bid lines should name the faceplate and the strike as items, and should state the rebate and door stop details the lock was matched against.
The join between the lock and the network. A networked door needs a controller, a network point, cabling, a power arrangement and a commissioning activity, and each has an owner in a new build: the network package, the electrical package, the main contractor's package, or the client's own systems team. Line supervision and the egress interface have to be coordinated against a door position and a cable route fixed in the wall design long before the hardware is ordered. Where the hardware package is silent on the join, the assumption is that somebody else has covered it, and the specification for that other package may not have mentioned hardware at all. Smart Lock Network Requirements covers what the network designer needs to be told.
The generalisation is worth carrying into the tender: name the joins. A tender that states who owns the bore, the closing edge, the controller and the credential records produces comparable bids. A tender that names none of them produces bids that are individually reasonable and collectively incompatible, and the incompatibility surfaces at first fix.
Three cells that travel as conditions, not as variables
Leveling normalizes, and normalization averages, and an averaged compliance answer satisfies nobody. Three cells in a door hardware tender are conditions, not variables: they belong in a schedule attached to the comparison rather than in the elements that get normalized.
| Cell | What it governs | What it does not answer | Who decides |
|---|---|---|---|
| Egress and free egress on the escape route | That the escape route operates without special knowledge, and how the door is released by panic hardware | Whether a locking arrangement is acceptable on a given door | The fire authority, with local building control |
| The fire door assembly | The performance of the door as an assembly of leaf, frame, hardware and locks | Whether a component added after the test remains within that evidence | The fire authority |
| Operable parts and accessible approach | The force, motion and approach required of the operable part | Whether the door as installed complies, because only the installed door can be checked | The access consultant, with building control |
CONFIRM-WITH-FIRE-AUTHORITY — the escape route and any door forming part of a fire door assembly are settled by the fire authority for the project, and the consultant's question is which doors those are and whether the locking arrangement on the schedule is the arrangement that was approved. The authority does not answer by general rule; it answers on the project's own drawings, so the evidence needed is the approved drawing and the written approval, dated and naming the door.
CONFIRM-WITH-LOCAL-BUILDING-CONTROL — building control signs off on the building as a whole, and it is where the interaction between the locking arrangement, the door and the escape route is settled for the local jurisdiction. The consultant's task is to carry that interaction in the schedule and to obtain the outcome in writing before the door order is released.
CONFIRM-WITH-ACCESS-CONSULTANT — an operable part is verified on the installed door, not on the sample and not on a datasheet. The access consultant who sets the project's expectations should be given the door set references and the intended operable part, and their requirement then becomes a condition of the award rather than a preference to be weighed against other elements.
The check that belongs to the specifier here is the third of the standing five: is the locking arrangement for the escape route the one the authority approved, and is that approval written down and dated? A condition cell without a dated approval is not a condition; it is an open item with a reassuring label, and it will surface during inspection with no time left to change anything. These three cells therefore travel in a separate schedule with a status column — approved, not approved, not applicable — and each line names the authority that has to sign it and the date the answer is needed by.
Documents that turn an assertion into a checkable claim
A bid that says a product complies is not an input to a comparison. It is an assertion that will have to be re-opened later, at the cost of whoever holds the project. The bid that levels arrives with documents a reviewer can check without picking up the telephone.
| Document in the bid | What it must show | Who holds the original |
|---|---|---|
| Dimension drawing for the exact variant | The dimensions of the variant offered, dated | The bidder |
| Cut list or template | The preparation the price assumes | The bidder |
| Evidence for each regulated characteristic | A named document, a scope and a holder, not the word "complies" | The bidder |
| Assumption register | Every assumption the price rests on, written down | The bidder |
| Exclusion list | What the price deliberately leaves out | The bidder |
| Platform and credential description | The management path, the credential methods, where the records live | The bidder |
| Service and spares arrangement | Who answers, who replaces, what is stocked | The bidder |
The row doing the most work is the third. The standing test for evidence is whether it names a document, a scope and a holder. A statement that a product meets a requirement, without the document that demonstrates it and without a named party who stands behind it, transfers the whole problem to the client at handover. The supplier's obligation is to evidence the requirement; the specifier's obligation is to write that into the submittal requirement rather than to make the determination themselves.
The platform row needs care on a new-build scheme, where one project can carry several credential approaches and a single bid can hide a mixed proposition. Ecosystem names and credential methods appear on marketing pages, and in a bid they belong as a question with a date attached: confirm on the dated product page for the specific SKU, and confirm the configuration that will actually be ordered rather than the one illustrated. A capability claim is CONFIRM-PER-SUPPLIER — the supplier confirms it for the specific variant on the dated page, and no field is extended from one variant to another. Where a record is empty, the model does not offer that capability, and the bid says so rather than leaving it implied.
Provenance is the other half of the evidence question. The documents in a bid should come from the bidder, with a named document and a named holder, and the consultant should know how to check the party behind them: Smart Lock Supplier Verification China and Smart Lock Factory Audit China set out what that check involves, and Smart Lock RFQ Guide for International Buyers covers the version where the bidders are not in the same jurisdiction. Two cluster articles approach the same work from the other two ends: Bid Leveling: Documents a Specifier Should Demand from the tendering side, and Bid Leveling: Template Language for a Specification from the specification side. The constraint this article adds is the one they both inherit: the door is not there yet, so the documents have to describe a predicted door rather than an observed one.
The moment the prediction meets a real leaf
A sample and a mock-up evidence fit. They do not evidence performance, and on a new build they are the first point where the schedule's prediction meets a real object. That makes the timing of the sample a commercial decision rather than a formality, because the sample has to be fitted to a leaf from the same release as the hardware award if the fit question is to close on the project rather than in a showroom. Where the door order is released later, the sample is fitted to a factory mock door, the fit question stays open, and the first real test of the schedule happens on site with the trades already working. That outcome is avoidable, and it is decided by the award date rather than by the tender date.
The same timing applies to the finish, and the finish is the cell where the new-build context is most often mishandled. A photograph shows the finish; it does not specify it. The RAL or colour reference is the identifier, and a bid that names a finish only by photograph has given the factory a range rather than a target. The finish must be named in the bid with its reference and its application, and the sample then evidences the match between what was ordered and what was agreed.
The procedural rule that keeps this honest is the one the whole submittal process runs on: the submittal is not approved until the review is recorded and dated. A sample approval on a new build is an approval of fit and finish against a named door set and a named finish reference, recorded with a date and a name. It is not an approval of the product in general, and it is not a performance approval. The mechanics are handled in Smart Lock Surface Finishes and in What Should I Check Before Approving a Smart Lock Sample?.
Nine gates, and the one that closes the window
Here is the point that distinguishes new-build leveling from other kinds of bid comparison. The leveling window is not the tender window. From the issue of the tender to the award, the bid is a document that can be amended. From the release of the door order onwards it is a commitment against a leaf being cut and a frame already built into the wall, and a scope difference becomes a variation against an item in manufacture.
| Gate | What it settles | What must exist to pass it |
|---|---|---|
| Tender issue | The door set references and the elements to be priced separately | A schedule with no silent cells across the eight elements |
| Bid receipt | That every bidder has answered on the same form | A declaration of any departure from the requested basis |
| Basis check | Whether each bid can be read on one basis | The eight elements answered, no assumption left implicit |
| Clarification | The silent cells, answered in writing | Questions sent to every bidder, not only the low one |
| Normalization | One declared basis per element | A written basis the client has seen |
| Comparison and recommendation | Which bid is recommended, and on what | The comparison with exclusions and open items |
| Award and door order release | The commitment | The signed award, the leveling record, the conditions |
| Sample and first article | Fit and finish against a real leaf | The sample approved, recorded and dated |
| Handover | The record the client inherits | The as-built pack, credentials transferred, spares |
The ordering matters. The basis check precedes the clarification, and the clarification precedes the normalization, because a normalization performed before the silent cells are answered is a second guess dressed as a method. A gate passed without its condition does not become a gate passed; it becomes a risk carried forward into the next stage, where it is more expensive to carry.
flowchart TD
A["Tender issued against one door schedule revision"] --> B["Bids returned in each bidder's own quotation format"]
B --> C["Basis check on scope, geometry, system scope and documents"]
C --> D{"Is any schedule cell silent?"}
D -->|"Yes"| E["Ask every bidder to state the reading they priced"]
E --> F["Amend the schedule cell and reissue it to every bidder"]
F --> C
D -->|"No"| G["Normalize the eight elements onto one declared basis"]
G --> H["Compare door set by door set against the door order"]
H --> I["Carry egress, fire and access as conditions with a status"]
I --> J["Recommendation with named exclusions and dated open items"]
J --> K["Award signed and the leveling record issued with it"]
K --> L["Door order released and the leveling window closes"]
The clarification request goes to every bidder, not only to the one whose bid looks unusual. If one bidder is asked to re-price and another is not, the result is not a comparison and cannot be defended when it is questioned. Where a clarification changes the basis for everyone, the tender is reissued with the amended cell and every bidder prices against the same amended document.
Who owns the record is the other half of the gate discipline. The leveling record belongs to the consultant who ran the exercise and is handed to the client with the award, dated, so the client holds a documented basis for a building it will operate for. A record nobody owns cannot be relied on when the first variation appears. The programme responsibilities around this work are set out in Smart Lock Contractor Procurement Guide.
How a door set changes state once it has been ordered
The same idea can be read as a short state machine, which is useful when a variation has to be classified rather than argued about.
stateDiagram-v2
[*] --> Predicted
Predicted --> Revised: schedule cell amended and reissued
Revised --> Predicted: bidders confirm the amended basis
Predicted --> Ordered: door order released
Ordered --> Cut: leaf manufactured and frame set
Cut --> Verified: sample checked against a real leaf
Verified --> Handed: records and credentials transferred
Handed --> [*]
Ordered --> Variation: scope change after release
Variation --> Ordered: variation priced and accepted
A change to a predicted door set is an amendment and costs a reissue. A change to an ordered door set is a variation, and the party that ordered it holds the consequence. That is why the leveling window has to close at the door order rather than at the award recommendation: after that point the same words describe two different things.
The sheet the client keeps after the award
The output of a leveling exercise is not a comparison table in a meeting. It is a sheet the client keeps, because it answers the questions that arrive two years later: why this lock on this door, who agreed to cut the bore here, and what was excluded.
| Element | Basis stated at the award | Where the answer came from | Excluded by agreement | Still open at the award |
|---|---|---|---|---|
| Lock supply | ||||
| Door and frame preparation | ||||
| Closing edge and frame hardware | ||||
| Network and power | ||||
| Software and credentials | ||||
| Commissioning, programming and training | ||||
| Documentation | ||||
| Service and spares | ||||
| Egress condition | Not a cost element | |||
| Fire assembly condition | Not a cost element | |||
| Access condition | Not a cost element |
Four rules govern the sheet, and they are the difference between a record that works and a document that gets filed.
No single figure. The sheet does not reduce to a number, and the comparison it supports is not a ranking. A client who wants a single figure is asking for the one thing the exercise cannot produce, because it would be the total of a scope nobody can describe.
A basis per element, not a summary. Each line says what was priced and on what basis, which is what lets a later reader separate a genuine change in requirement from a misunderstanding that was never corrected.
The exclusions are named. Where the client and the consultant agreed that an element was out of scope, the sheet says so and says who agreed. Silence about an exclusion becomes an assumption about entitlement, usually at the worst possible moment.
The open items are dated. An item left open at the award carries a date by which it closes and a name against whom the answer is required. An open item with no date becomes a dispute with no deadline attached to it.
Two mistakes are specific to new construction, and both are prevented by the sheet rather than by a second review. The first is reading a scope difference as a price difference: the lower bid is attractive, and the reviewer starts hunting for a reason to prefer the other bid, when the reason is usually hiding in an element the lower bid did not price. The reviewer's task is to find the assumption that separates the bids and let the client trade it knowingly. The second is comparing catalogue geometry against scheduled geometry, which surfaces as a cut in the wrong place at first fix.
There is a fifth rule, and it belongs to the life-cycle rather than to the leveling. The last of the standing checks asks what the client assumes about the life of the installation, and it insists that the assumption is written as an assumption, with the figure the client supplied rather than a figure supplied by anybody else. The consultant's job is to record the client's own assumption, attributed, and to make clear which parts of the model are assumptions and which are committed scope. The commercial consequences of how that model is built are treated in Bid Leveling: Cost and Risk Consequences, and the mechanics of running a multi-supplier comparison are in Smart Lock Quotation Comparison.
Ten questions to send to every bidder in one request
These questions are what the leveling exercise is made of. Each is answerable in writing, each has a place in the sheet, and they can be sent unchanged to every bidder, so the answers are comparable.
- Which door set references and door number ranges does this bid cover, and which references are excluded?
- Which door schedule revision, and which date, did you price against?
- For each door type offered, which leaf thickness, handing and rebate did you assume, and from which drawing?
- Which cut list did you price against, who performs the bore, and what happens to the door if the leaf arrives unprepared?
- Are the faceplate, strike and cylinder included, and which door manufacturer data were they matched to?
- Which of the eight elements are inside this bid and which are outside it? If an element is outside it, name the party you assume will provide it.
- Which doors are networked, which standalone, and what does the network scope include — controller, network point, cabling, backup, commissioning?
- What does the client own at the end of the contract, what is licensed, what is hosted, and how are credential records handed over?
- Who enrols the first credentials, who trains the operators, and what is tested and recorded before handover?
- Which assumptions does the price rest on, and which exclusions does it make?
Question 6 carries the most weight and is the one most often left out, because a bidder who has not priced an element has no ready answer and may prefer silence. Silence is recorded on the sheet as an exclusion with a named party, which is both accurate and useful.
The answers belong in the bid. An answer given in a meeting and not confirmed in writing is not part of the basis. Where clarification is needed after the clarification window has closed, the answer is added to the leveling record with its date and the name of the person who gave it, and if the answer changes the basis for anyone it changes it for everyone. The comparison that follows is a mechanical step, and the mechanics are set out in How Do I Compare Smart Lock Factory Quotations? and in Bid Leveling: Reviewer Checklist.
Who holds the record after handover
The fourth of the standing checks asks who owns the record after handover and in what form the handover actually happens. On a new build the answer is easy to leave open, because the building is handed over in stages and the hardware is often commissioned before the final handover event. Three categories of record have to be named, each with an owner, a format and a date: the technical record of drawings, schedule revisions and test results; the operational record of credentials, key control and spares; and the authority record of dated approvals filed against the door set references they apply to. The handover pack is the last occasion on which anyone can insist that the configuration, the credentials and the approvals travel together, and the wider procurement checks around it are collected in Smart Lock Procurement Checklist for B2B Buyers.
What this method cannot settle, and where the figures come
Leveling normalizes scope onto one declared basis. It does not decide the design, and it does not replace the authorities that hold the design accountable. Where a standard or a local requirement is in play, the question to ask is what the document governs and what it does not answer, and the evidence to demand is a named document with a stated scope and a named holder. The determination itself belongs to the authority that holds it, and the specifier's job is to write the evidence into the submittal requirement rather than to make the determination.
CONFIRM-WITH-LOCAL-LEGAL — bid differential conditions, retainage, bonding, and any term that makes the award conditional on a later decision are matters for the client's own counsel. This article does not set them, does not describe how they are applied, and does not tell a consultant how to weigh them. It stops at the point where the commercial consequence becomes a legal one.
This article carries no price, no bid differential, no percentage and no discount figure, and that is a deliberate position rather than an omission. A figure without a declared basis is precisely the thing the exercise exists to prevent, and a published figure would be a bid difference with no scope attached to it. Every project-specific figure comes from a quotation raised against the leveled basis — the element list, the door set references, the schedule revision and the date. Request a quote is the route for that, and the enquiry that comes back should carry the same basis the leveling sheet carries.
The last thing the consultant should do, and the thing most often skipped under programme pressure, is write the method down. A short note explaining what was compared, on what basis and what was excluded costs an afternoon, and it is the difference between a tender the client can administer and one that has to be re-argued every time something changes. The next reader of this project will not have the consultant in the room, and the sheet is the only thing that will speak for them.

