A PDF is frozen the second you export it, and every step between the customer and the price is a place the sale stops.

Because the quote has to survive being forwarded to a spouse, opened in a truck, and read again three weeks later — and a file does none of that well. Why they get in with no password, and how the same address carries them on to the build once the job starts, is set out on one customer link, no password.

A big number from an unknown company reads as a risk, and the customer's first instinct is to go and get two more quotes.

A figure is easier to accept from somebody who looks established, and the cheapest place to look established is above the figure.

A line reading 40×64×16 asks the customer to imagine it, and half of them imagine something else.

The moment they turn it around, the conversation stops being about a number on a page and starts being about their building.

Customers nod at a drawing they cannot read, then discover at framing that they thought the doors were somewhere else.

Everything they are buying should be readable by somebody who has never read a construction drawing — because that is who is signing. Where the design does not answer a row, the sentence that fills it is your wording, per company, rather than post-frame boilerplate on a stud-frame job.

The usual version is three phone calls, a colour chart photographed badly, and a wrong order.

It is the section customers spend longest on, and it settles the colour conversation before it becomes a change order — which is the expensive way to have it.

A line item reading “Concrete Pad — $17,400” invites exactly one question, and it is not a good one.

Two sections, two jobs: this one answers what am I getting, and the price block answers what does it cost. Asked and answered separately, neither has to carry the other's weight.

One level of detail for every customer means either over-explaining to the ones who don't care or under-explaining to the ones who do.

A builder who bundles labour into a rate needs the unit price hidden. A builder quoting a package needs both hidden. Neither should have to keep a second version of the quote to get it.

An extra the customer might want either goes in the price and makes the whole quote look expensive, or goes in a second quote they have to ask you for.

People like choosing more than being sold to. A headline price without the extras feels affordable, and adding the approach themselves feels like their idea — which is exactly how an upsell sticks. For you it's a bigger job without a revision.

One total at the bottom of a PDF invites exactly one question, and it is “why is it so much?”

A customer reading a breakdown that adds up is evaluating a build. One who spots a gap between the lines and the total goes hunting for your margin, and starts to haggle. Built-in prices keep them on the building.

The money conversation gets had on the day you need the deposit — the day you have least leverage and they have most doubt.

In writing, before commitment, is the only comfortable time for either side to have it.

Some customers genuinely cannot evaluate $68,400 — it is not a number that appears anywhere else in their life.

Nearly everyone can evaluate a payment, and a buyer who can evaluate the number can make a decision instead of postponing one.

The same five questions arrive by phone, one at a time, usually on a Monday.

They are answered at two in the morning when the customer is actually thinking about it — which is when the decision is really being made.

Buried in the contract, the warranty only gets read after the customer has already decided.

It is worth more here, where it is still helping them decide, than in the document they sign afterwards.

The gap between “yes” and a signed document is measured in days, and it is where deals go cold.

The signature is captured at the moment of most conviction — the end of the argument, not a week after it.

A customer who decided at section four has to scroll past eight more sections to act on it, and some of them don't.

Never make somebody who has already decided go looking for the way to say so.

An email saying “looks good” is not an approval, and you have to keep it forever in case anyone asks.

What was approved, by whom, at what price and at what time, on one artefact — because that is the question that gets asked when something goes wrong.

Three PDFs called final, final-2 and final-USE-THIS, and the customer signs whichever one they happen to open.

Exactly one version is live, so the version they can reach is the only version they can sign. Each one is a dated snapshot rather than a figure that keeps recalculating behind you, and the design becomes the estimate sets out why that freeze is deliberate.

“Why is it four thousand more than last time?” arrives on the phone mid-sentence and gets answered from memory by whoever picked up, so the customer decides the price moves depending on who they ask.

Colour is doing real work rather than decorating: with two totals of the same job on one screen, attaching a figure to the wrong version is the only mistake still available, and the colour takes it away.

The customer forwards themselves the estimate in March, approves it in May from the March email, and the builder — who revised it twice in April — books the job at the current figure. Nobody finds out until the first invoice does not match what the customer thinks they signed.

Because this is the failure that costs real money and gives no warning: everybody believes the deal is agreed, and the number the customer agreed to is one revision old. Catching it at approval is a two-minute conversation; catching it at invoicing is a renegotiation.

Working out what a change is worth means editing the live estimate, which the customer might open at any moment. So the builder either does the sum on paper and quotes it roughly, or edits the real thing and hopes nobody refreshes.

Because the useful moment to ask “what if we dropped the wainscot” is while you are on the phone with them, and every tool that makes you commit before you can see the number turns that into a callback. Separating draft from live is what lets you answer immediately without risking the version they are looking at.

Silence after sending a quote is ambiguous, and the two things it can mean need opposite follow-ups.

A customer who has read it four times and one who has never opened it are completely different phone calls. Without this you make the same call to both and get it wrong half the time.

A six-figure total is a number most buyers cannot place against anything else in their life, so the decision gets parked until a lender conversation they have not booked yet.

It is a real amortisation of the real figure rather than a rounded teaser, because the number gets repeated to a bank and has to survive being checked there. Putting it directly beneath the total, rather than in its own section, keeps it reading as the same price said another way.

Your lender does 6.9% over 84 months, the quote advertises 8.5% over 60, and the payment your customer reads is worse than the one you could actually get them.

The rate is stored as a decimal rather than a percent, in a column that will not hold anything at ten or above — so a percentage pasted in raw is refused outright instead of being printed on a customer's quote as a three-figure APR. Set nothing and the page still falls back to 8.5% over 60 months, which is what every account used to get whether it fitted them or not.

The customer repeats “they said about two thousand a month” to their bank, gets quoted more, and comes back believing you sold them a rate you never had.

A monthly figure said out loud by a rep is where a complaint starts. Written under the number with its assumptions attached, it is a statement the customer can re-read — which is the whole reason it is not in a footer nobody scrolls to.

A cash buyer reads a monthly payment under a price he was going to write one cheque for, and the first thing he asks about is interest he was never going to pay.

It is its own row under Your Investment rather than a child of the total, because showing the price and offering a payment are two different decisions. With the total hidden the box takes its own divider instead of floating loose on the page.

You retype the same warranty paragraph into every quote, and by spring the copy on a job you sold in October says something different from the one you sent this morning.

Set once and overridable per job, so an unusual build can carry different terms without anyone touching the version every other customer reads.

Answered out loud, those five come out slightly different every time, and when the customer remembers the permit conversation differently in week three there is nothing to point back at.

They ship pre-written so a new builder's quote is never a blank page, and each can be re-cut for a single job, so a longer build says so on its own document instead of in a phone call nobody wrote down. A written answer under the price is a promise the customer can re-read; a spoken one is not.

Nothing on your quote says you are licensed, insured and eleven years deep, so a buyer holding three pieces of paper has only the price to compare them on.

The claims are picked from a prepared list of twenty rather than typed into a box, which is what keeps the counts that should move wired to your own record while the fixed ones stay fixed. The completed count rounds down to the nearest five, because a count worked out to the unit reads as a boast; hovering pauses the rail, and a visitor who has asked for less motion gets a still layout instead.

You change your business line and the old one is still sitting on live estimate pages, on dashboard call buttons and on every quote you sent last month, each holding its own copy.

Backup chains are how a quote ends up printing a personal number somebody replaced two years ago — or, once, the customer's own line, telling them to ring themselves. Rendering nothing when the field is blank is the deliberate call: a missing number reads as an oversight, a wrong one reads as a builder who does not answer.