The spreadsheet arrives on a Tuesday. Column A is your material. Column B is the contractor's alternative. Column C is the saving. Nobody has asked you anything; they have simply listed your building's surfaces in descending order of how much money can be taken out of them, and the handmade brick is at the top because it always is. You have two days to respond. The client has already seen column C.
That meeting is where a materials palette either holds or evaporates, and the outcome was decided months earlier — back at Stage 2, when the board was assembled. Not by how good the materials were. By whether anyone had asked them the boring questions.
Here is the uncomfortable diagnosis. What gets pinned up at concept stage is usually a mood: nine photographs of other people's finished buildings, a fan of colour chips, a scan of some oiled oak. Beautiful. Coherent. Completely untethered from where any of it comes from, who makes it, what it costs per square metre, how long it takes to arrive, and who on a building site is competent to install it.
That isn't a palette. It's an aspiration with a nice layout.
A palette is a set of decisions, each with a reason behind it and an answer ready for the four questions somebody is going to ask. The board is just where you keep them. When a QS asks why the soffit is fibre cement and not plasterboard, "because it looked right" is not an answer, it's a surrender. And the thing about value engineering is that it doesn't attack your worst decisions — it attacks your least defended ones. Whatever you cannot explain, goes.
So the work is to make each material defensible before anyone attacks it. Four filters do most of that work.
Filter One: What Cost Band Is This In?
You do not need a price. You need a band, and you need it early, because the difference between a decision you can afford to take and one you can't is usually an order of magnitude, not a percentage.
In practice the useful question is not "what does this cost" but "what is this the expensive version of". Every material sits in a family. Handmade brick is the expensive version of a wirecut facing brick. Terrazzo is the expensive version of a polished screed. Solid brass ironmongery is the expensive version of stainless. If you know the family, you know where the pressure will come from, and you know the shape of the substitution before it's proposed.
Get the QS to band things for you at Stage 2, even informally. What we've found is that a cost consultant asked "is this a £ thing, a ££ thing or a £££ thing" will answer honestly and quickly, whereas one asked for a rate for a material you haven't specified yet will either refuse or invent something cautious and high. Ask the question you can actually get answered.
And be deliberate about where the money goes. A palette where everything is mid-range reads as timid and gets cut anyway. A palette with three genuinely expensive moves and a lot of honest, cheap background is easier to defend, because the expensive moves have a job. "That's the only stone in the building, it's at the entrance, and it's the thing people touch" is an argument. "Stone throughout" is a target.
Filter Two: How Long Does It Take to Arrive?
Lead time kills more specifications than cost does, and it kills them later, when there is less room to recover. A material that is affordable and unavailable is not a material.
The discipline is simple and almost nobody does it at concept stage: ring the supplier. Not the website. The supplier. Ask what the current lead time is on the specific colour, the specific format, the specific finish, and ask whether that changes if the order is placed in, say, February rather than September. Write the number they give you down on the board next to the material, with the date you asked, because that number ages.
Then compare it against the programme. If your bespoke terracotta baguettes need ordering before the contractor has been appointed, you have not specified a material — you have specified a procurement problem, and someone else will solve it by choosing something else.
Certain categories are structurally fragile here. Anything imported, anything made to order, anything with a single manufacturer, anything in a colour that isn't in the standard range, anything that needs a mould. Anything where the word "bespoke" appears. That doesn't mean don't use them — it means use them where you have decided the risk is worth it, and know in advance what happens if the answer comes back as thirty weeks.
Filter Three: Who Actually Installs It?
This is the filter that separates people who have built something from people who have only drawn it. A material is not an object, it's a trade. Specifying it means specifying a subcontractor, a skill level and a sequence.
Ask of every material: is this something a general contractor's own team can do, or does it bring a specialist onto site? Specialists mean lead time, preliminaries, coordination and a price that moves depending on how busy that specialist is. Two trades meeting at a junction is where the detail either works or leaks — and where the contractor's price goes up, because nobody wants to own an interface.
We teach this because it changes the design, not just the schedule. A palette assembled from materials that are locally available and locally installable tends to be cheaper, faster, better executed and — this is the part people miss — more likely to survive tender intact, because there's no obvious alternative sitting next to it on the spreadsheet offering the same look for less. The cheap version and the good version are the same version.
Filter Four: What Does It Look Like in Five Years?
Almost every reference image is of a building about four weeks old. Photographed in the morning. Before anyone lived in it.
So run every material forward. Where does the water go, and what does it leave behind. What happens where hands touch it. What does it do where the sun hits it and where it never does. Whether it weathers to something better or just to something dirty. Whether a cleaner will use the wrong product on it, because they will.
This is also, incidentally, your strongest argument in a VE meeting, and the one clients respond to most. Cost consultants are comparing capital cost; you are the only person in the room talking about the next twenty years. The oak handrail versus the powder-coated one isn't a taste argument, it's a maintenance argument. Make it that.
Two Choices for Every Element, and Both of Them Yours
Here is the practical mechanism that makes the whole thing hold. For every element of the building — external wall, roof, window, internal wall finish, floor, ceiling, ironmongery, sanitaryware — carry two materials. A first choice, and a second choice you would genuinely be content to build.
Not a fallback you despise. A real one. If the second choice is something you'd be ashamed of, you haven't done the work; you've just left a gap for someone else to fill.
The reason is tactical. When the VE list arrives, the architect with one option is negotiating from a position of loss — every conversation is about what's being taken away. The architect with two is negotiating from a position of choice. "If we can't have the handmade brick, the answer isn't the contractor's reconstituted stone, it's this wirecut in a closer tone with a recessed perp — here's the saving, here's what it costs us in quality, and here's the thing we are not giving up in exchange."
You have just done the QS's job better than the QS, and you've kept control of the substitution. That is what defending a palette actually looks like. It is not saying no.
Write Down Why, Not Just What
Every material on the board needs one line of reasoning attached to it. One sentence. What is this doing that nothing else does.
This is the bit that converts a board into an argument, and it's the same discipline as researching precedents so they actually change the design — the value isn't in the image, it's in the extracted reason. "Zinc standing seam because the roof is visible from the street above and the seam lines follow the fall" survives a meeting. A photograph of zinc does not.
Those lines also do something useful months later, when a substitution is proposed and you're three projects deep and can't remember why you chose the thing. The reason tells you immediately whether the alternative satisfies it. Half the time it does, and you should say yes fast and bank the goodwill. The other half it doesn't, and now you can say exactly why — which is the only version of "no" that anybody listens to.
Keep all of it in one place: the image, the supplier, the cost band, the lead time and the date you got it, the trade, the ageing note, the second choice, the reason. One board, exported, sent to the client and the QS at Stage 2 rather than discovered by them at Stage 4.
Documentation Is Where Ambiguity Gets Monetised
One last thing, and it's the one that costs practices the most money. Vague specification is an invitation.
"Timber cladding" priced by a contractor is the cheapest timber cladding that contractor can source. "Or equal approved" means approved by whom, against what, and if the answer isn't written down, the argument happens on site at the worst possible moment. Name the product, the manufacturer, the format, the finish, the fixing. Where you genuinely want to allow alternatives, say what an alternative has to match — the dimension, the tone, the durability class — so the test is yours rather than the contractor's.
The palette you can defend is not the one with the best images. It's the one where every material has been asked the four questions, has a partner waiting in the wings, and carries a single sentence explaining what it's for.
Assemble it that way and the Tuesday spreadsheet stops being an ambush. It becomes a negotiation you're prepared for, which is the most any of us get.
