Quick answer
Cut variations and RFIs by locking the spec before fabrication, routing all queries through one point of contact, and doing an on-site measure check before manufacturing. MTN Kitchens uses this supply-and-install approach across Auckland volume packages to keep developer and builder programmes on track.
Every variation and RFI on a kitchen package is margin walking off your job. A late dimension query parks the joiner. A bench colour nobody locked gets installed wrong, then re-cut. The handover meeting turns into a callback list. On one house it's a nuisance. Across a whole subdivision it's a programme problem, and the cost lands on you, not the supplier.
One source of truth, not four.
Key points
- Kitchen variations are rarely about kitchens. They are about ambiguity: specs never written down, measurements off a concept plan, and three people answering the same site question three ways.
- A signed spec per typology answers nearly every site question before it is asked, so the installer is never guessing the handle, the finish or the stone colour.
- Supply and install under one contract deletes the gaps between suppliers. One mob measures, builds, delivers and fits, so the benchtop templates off the real cabinets, not a drawing.
- A site check measure after framing and linings turns a future variation into a quiet pencil adjustment. On repeat jobs, prototype the first unit of each type before releasing the rest.
- The real saving on a volume package is not in shaving the carcass. It is in driving re-makes, re-fabrications and return visits toward zero.
Here's the thing most PMs work out the hard way: kitchen variations are rarely about kitchens. They're about ambiguity. Specs that were never written down, measurements taken off a concept plan, and three people answering the same site question three different ways. Sort those out and the noise stops. What follows is how we run volume packages out of our East Tamaki workshop to keep RFIs off your desk, and what you can ask any kitchen supplier to do the same. Most of it is process, not magic. The yards that do it well are just disciplined about the front end of the job, where every dollar saved is worth ten saved on install day.
Where the variations actually start
On volume work the culprits repeat. Cabinet sizes assumed off a drawing instead of measured against the GIB. A benchtop edge that got talked about on a site walk but never made it onto a sheet. An appliance spec that changed after the kitchen was drawn, so the rangehood cutout no longer fits. Waste, water and power landing a few hundred millimetres off where the joinery expected them. None of it is exotic. It's the same five or six gaps showing up on every unit.
And because the units repeat, one gap multiplies. A single vague line in the spec turns into an RFI on every unit. Or worse, quiet assumptions nobody flagged, which then surface as defects on handover day when you've got least time to deal with them. Resolve the ambiguity once, at the front of the job, and you're not firefighting it unit by unit under programme pressure. This is the same discipline that lets you keep kitchen spec consistent across 30+ units without a different fix on every floor.
It helps to be honest about who creates the gaps, because it isn't always the kitchen supplier. The architect documents to concept level and leaves the detail to the trades. The developer changes the appliance package after pricing to chase a margin. The builder moves a waste during first fix because the slab penetration was out and never tells anyone. Each decision is reasonable on its own. Strung together with nobody holding the thread, they land as a variation on the kitchen because the kitchen is the last thing in and the most visible thing wrong.
A locked spec answers the same question on every unit
Before anything is cut, we want a signed specification per kitchen type. For a developer that's a schedule naming exactly what we supply and install: carcass, door finish (2-pac lacquer, melteca, timber), benchtop material and edge profile from laminate up to engineered and natural stone, hardware, runners, splashback, and an appliance schedule with model numbers against it. If there are three apartment typologies, each gets its own sheet and every unit points at one of them. If you're weighing up the finishes that go on those sheets, our breakdown of lacquer, melamine and timber cabinet finishes is a useful companion.
Once that's signed off, nearly any site question already has an answer on paper. The installer isn't guessing the handle or the stone colour. It's named. That's also where building the same typology to the same drawing pays off. Unit 3 and unit 33 come off the bench identical, so your defect list at handover is short and the body corporate isn't chasing you six months later over a kitchen that doesn't match its neighbour.
What "locked" actually means
Locked doesn't mean nothing can ever change. It means changes go through a known door instead of getting whispered to a chippie on site. A locked spec has a version number and a sign-off date. When the developer swaps the 600mm oven for a 900mm one, that's a documented variation with a price and a programme impact attached, not a surprise the installer finds when the cabinet won't take the appliance. The discipline isn't about being rigid. It's about making sure every change is visible, costed and agreed before it touches the saw.
One contract, one contact, one source of truth
RFIs breed when nobody owns the interface. Cabinetry from one supplier, benchtop from a second, install by a third, and every seam between them becomes a question you chase. Who measured? Who owns the cutout that doesn't line up? The PM ends up refereeing trades instead of running the build.
Supply and install under one contract deletes that whole category. One mob measures, builds, delivers and fits. There's no daylight between the drawing and the person fixing it, and you get one invoice instead of a reconciliation. Ring one number and you get a straight answer, not a query bouncing between two yards in Manukau and East Tamaki for a week. When the benchtop fabricator and the cabinet maker are the same company, the template comes off the actual installed cabinets, not a drawing, and the stone fits. That single fact kills more benchtop variations than anything else on a job. There's more on this in why builders outsource kitchens and where the line sits.
It also changes who carries the risk. Split the package across three suppliers and the gaps between them are your problem by default, because no one signed up to own them. Put it under one supply-and-install contract and the measure, the build and the fit are one party's responsibility. If the cabinet meets the wall 30mm out, that's their re-make, not your variation. You want that line drawn before the job starts, not argued about at handover.
Measure on site before the saw runs
The dimensional RFIs are the expensive ones, and they're almost all avoidable. Cabinetry built off a concept plan that then meets a wall 30mm out is a re-make, on someone's dollar, and a return trip to site. Cabinetry check-measured on site, with framing up and services confirmed, fits first time. We go deeper on this trade-off in site measure vs plan measure.
It comes down to timing. A check measure once the framing and services are in, before the cut list releases, turns a future variation into a quiet pencil adjustment. It catches what the drawings miss: out-of-square walls, a waste that moved during first fix, a window sill sitting lower than documented, a meter box that ate 40mm of the run. Catch it before the panel saw runs and it costs nothing. Catch it on install day and it's a return visit, a programme slip and a variation note you have to explain upstream.
On a repeat-typology job there's a smart middle path. Measure the first unit of each type fully, build it, install it, and treat it as a prototype. Walk it with the builder and the developer, confirm the fit and the finish in the flesh, then release the rest of that typology to manufacture. You spend a few days up front and you save yourself the same mistake replicated across every unit. On a single villa reno there's no prototype to lean on, so the site measure has to be exact the first time, which is why those jobs reward a measurer who's seen a hundred old Ponsonby kitchens and knows where the floors slope.
A worked example: 24 townhouses in Flat Bush
Say you're running a 24-unit terraced housing development in Flat Bush: three typologies, eight of each. Two-bed units get a galley kitchen, three-bed units get an L-shape with an island, and the corner units get a scullery off the main kitchen. If you're still settling on those layouts, kitchen layouts explained walks through where each one earns its keep. The temptation is to start the kitchens when the first units are watertight and just keep feeding the programme. That's how you end up with 24 slightly different kitchens and an RFI a day.
Here's the sequence that keeps it dull. At pricing stage we get one signed spec sheet per typology with appliance model numbers locked, because the oven and rangehood sizes drive the cabinetry and the cutouts. While the first units frame up, the drawings get done off the consented plans, three sets, not 24. Once typology A's first unit is lined and the services are in, our installer check-measures it against the joinery layout, confirms the waste and power, and that becomes the control measure for all eight A units. We build and install that first one as a prototype, walk it with the PM, and only then release the remaining seven to manufacture.
Typologies B and C follow the same rhythm, staggered behind A so the workshop runs flat out without the site getting jammed. Deliveries land against the construction programme, not in one drop that clogs a unit you can't work in yet. By the time the last corner unit's scullery goes in, the crew has built that exact kitchen seven times and the install is a five-to-seven day single hit per stage, not a string of return visits. The defect list at handover is short because the prototypes caught the real-world problems before they multiplied. The developer sees one invoice per stage and the body corporate gets 24 kitchens that match.
What goes wrong, and what it costs you
The failures are predictable, which is the good news, because predictable means preventable. A few that turn up on Auckland jobs again and again:
- Appliances bought late or changed after the kitchen's drawn, so a 900mm cooktop meets an 800mm cabinet. Now you're re-cutting benchtops and re-making a base unit on install day.
- Benchtop templated off a drawing instead of the installed cabinets, so the stone fights the wall and you're chasing a re-fabrication with a two-week lead time sitting on your programme.
- No prototype on a repeat job, so a fault in unit one gets faithfully reproduced right through the typology before anyone notices.
- Services set out to the architectural plan, not the joinery layout, so the waste lands behind a drawer bank and the plumber has to come back.
- Split suppliers with no owned interface, so the seam that failed belongs to nobody and the fix stalls while you work out whose dollar it is.
- Verbal site instructions that never make a sheet, so the variation is real but undocumented, and you find out at the final claim.
Notice none of these are about the quality of the joinery. A beautiful kitchen built to the wrong dimension is still a re-make. The money you lose on a kitchen package is almost never in the cabinetry craft; it's in the coordination around it. That's the part worth being ruthless about. Many of these gaps are the same ones behind kitchen install delays, which is no coincidence: an RFI is just a delay that hasn't landed yet.
| Variation trigger | What it costs you | The fix that prevents it |
|---|---|---|
| Appliance spec changes after drawing | Re-cut benchtop, re-made base unit, multiplied across units | Lock appliance schedule with model numbers at pricing stage |
| Benchtop templated off a drawing | Re-fabrication, ~2-week lead time on your programme | Same company builds and templates off the installed cabinets |
| Cabinet built off a concept plan | Re-make and a return trip when it meets the wall 30mm out | Site check measure after framing and linings, before cut list |
| Fault reproduced across a repeat job | One mistake replicated across every unit of the typology | Prototype the first unit of each type, walk it, then release |
| Verbal site instruction, never written down | Undocumented variation surfaces at the final claim | One named contact each side, version-controlled spec |
Lock the spec before the saw runs.
The pre-start checklist we run
Before the first kitchen on a package goes into manufacture, we want these locked. Ask any supplier the same and you'll sort the disciplined ones from the rest in about five minutes:
- Signed spec per typology: finish, benchtop material and edge profile, hardware and runners named
- Final appliance schedule with model numbers, so cutouts and clearances are right first time
- Services confirmed on site (waste, water, power, gas) against the joinery layout, not just the architectural plan
- Site check measure done after framing and linings, before the cut list releases
- A prototype unit built and walked on any repeat-typology job before the rest go to manufacture
- One named point of contact each side for any question or change, with a version-controlled spec
- Delivery and install sequence tied to your construction programme and CCC dates
Nearly every variation I see on a volume job traces back to something nobody wrote down. Lock the spec, measure the site, build the first one and walk it, and the RFIs just stop landing in your inbox.
What it costs, and the levers that move it
Nobody can quote a kitchen package off a blog, and you should be wary of anyone who tries. What's honest is the shape of the pricing and the things that swing it. Trade and wholesale pricing on a volume package, with no showroom markup, sits well below what a retail kitchen company charges per kitchen, and the gap widens the more units you run, because the design and setup cost amortises across the typologies rather than getting paid for once per kitchen. If you want the full picture on that gap, trade vs retail kitchens lays out what developers actually pay. All numbers below are indicative and plus GST.
The biggest lever is the benchtop. Laminate is the floor; engineered stone steps it up a clear band; natural stone steps it again and adds lead time and templating cost. On a 24-unit job, choosing engineered over natural across the board can move the package materially without anyone living in those homes feeling short-changed. Our laminate vs stone benchtops piece weighs that call in detail. Door finish is the next lever: melamine and melteca are the volume-friendly choice, 2-pac lacquer costs more and adds finishing time, and timber sits up the top. Appliance allowances are usually the developer's call and sit outside the joinery, but locking them early is what stops them turning into variations.
Then there's the quiet lever nobody prices but everybody pays: variations and RFIs themselves. A re-made base cabinet, a re-fabricated benchtop, a second install visit, the joiner's day lost waiting on an answer. None of that is on the original quote and all of it lands on your job cost. Get the front end right and that whole line item trends toward zero. That's where the real saving on a volume package lives, not in shaving a few dollars off a carcass.
On programme is the part that pays
Cutting RFIs isn't only about cost. It's time. Every RFI you don't raise is a day the joiner doesn't lose waiting on an answer, and every locked spec is a delivery that turns up when your programme calls for it. A kitchen that lands on the day the build needs it, fits first time, and clears its part of the defect list quickly is worth more to a developer than a slightly cheaper kitchen that drags. Tying those deliveries to the build is its own discipline, covered in kitchen lead times and your construction programme.
Over 23 years and 2,000-plus kitchens, with capacity for 10-plus a week, the developers who keep coming back do it because the delivery is predictable and dull, in the good way. That includes head-contractor kitchen supply to Spencer Henshaw. We're Site Safe qualified too, so the health and safety side doesn't spin up variations of its own when you've got other trades on a live site. Predictable is the whole product. The kitchens are good, but plenty of yards make good kitchens. Fewer make them land on the right day, every unit, without a query.
Common questions
Frequently asked questions
How early do you actually need the appliance schedule?
Before the cabinetry is drawn, not after, because the appliance models drive cabinet widths, cutouts, clearances and rangehood ducting. The single most useful thing a developer can do to cut variations is lock the appliance package at pricing stage and treat any later change as a formal, costed variation.
Can you really measure on site and still hit a tight programme?
Yes, because the check measure is timed to slot into the build, once the framing and linings are up and the services are roughed in, a window that already exists in the programme. On repeat typologies we measure the first unit of each type, prototype it, and release the rest, so we are not re-measuring 24 identical units.
What's the advantage of supply and install over just buying the kitchens?
One owned interface and one invoice. When the same company measures, builds, delivers and fits, there is nobody to point at if the cabinet doesn't meet the wall, so a bad fit comes off their margin instead of turning into a variation on your job.
Does this approach work for a one-off, like a Ponsonby villa reno?
It does, with one change: there is no prototype to lean on, so the site measure has to be right the first time, which matters in an old villa where the floors slope and nothing is square. The locked spec, single point of contact and measure-before-cut discipline all still apply.
How do you handle a developer who changes the spec mid-job?
Through the door, not around it. A change to a version-controlled spec gets a price and a programme impact attached, agreed before it touches manufacture, so it stays a managed variation rather than an RFI you explain at the final claim.
Send us the job
If you've got a project coming up in Hobsonville Point, Flat Bush, Papakura or anywhere across the isthmus, the fastest way to see whether this stacks up for you is to put a real job in front of it. Send your unit count and typologies, or just your brief and drawings, and we'll work out the spec sheets and the sequence that keep the variations off your programme.
We'll come back inside 24 hours with trade-priced supply-and-install numbers and a spec you can lock before a single cabinet is cut. One contract, one contact, one invoice, built and installed out of our own East Tamaki workshop. The sooner the spec is locked, the fewer surprises land on you later, so the best time to send it is now, while the package is still on paper.