Go and look at the job properly. Price the parts you can see from your own rates. Then deal with the parts you cannot see in one of three ways — exclude them, put a stated allowance against them, or say you will price them once they are visible. Put the result in a document with a scope, an expiry date and your assumptions on it.
Look at the job, and write down five things
Most bad estimates are bad because of something that was visible on the day and nobody recorded. Photographs help, but they do not capture the five things that actually move a price.
- Measurements, in the units you price in. Square metres, linear metres, number of rooms, number of fittings.
- Access. How your materials get in and your waste gets out. A 900mm side gate and no vehicle access is a real cost, and it is the one most often discovered on the morning of day one.
- Condition. What you are working onto or into — the substrate, the existing pipework, the age of the wiring. This is usually where the unknowns live.
- What is in the way, and whether it is staying. Furniture, a shed, a shrub the customer loves.
- Who decides, and when they want it done. A quote sent to someone who has to consult a partner takes a week longer to be accepted, and knowing that stops you chasing on day three.
Example
A five-minute site note, before any pricing happens
Written on a phone in the customer's garden. Everything on it either changes the price or changes how the quote should be handled.
- Job — replace rotten deck, rear garden, roughly 22 m²
- Access — side gate 900mm, barrow only, no vehicle access
- Condition — sub-frame looks sound, two posts soft at the base, cannot see the rest until the boards are up
- In the way — shrub at the north corner, customer wants it kept
- Decision — both owners, contact is Ruth, wants it done before August
Price from your own numbers
Build the figure from the things you already know: your day rate, the materials at what they cost you this month, the disposal, and however long the job realistically takes including the part where you are waiting for something to dry.
Then do the check that catches most errors — hold the total up against a job you actually did. If a deck you built in March came to $5,800 and this one is coming out at $3,900, one of the two is wrong and it is worth ten minutes finding out which. Estimating drifts quietly, and your own past invoices are the only correction that is specific to you.
Add contingency inside your prices rather than as a line called contingency. A visible padding line invites negotiation about the padding.
Three honest ways to handle an unknown
Every job has something nobody can see yet. The mistake is not failing to predict it; it is pretending it does not exist until the invoice.
Exclude it
The cleanest option when the work is genuinely separable. “Excludes any repair to the sub-frame beyond the two posts noted.” The customer knows it might come, and knows it is not in this number.
Allow for it
A stated allowance — a provisional sum — for a specific possibility, charged only if it happens. Use it when the thing is likely enough that excluding it would make your total misleading. It keeps the headline figure honest, which matters when the customer is comparing quotes.
Refuse to price it yet
The right answer when the range is genuinely enormous. “If the joists are gone we are into a different job, and I will price that when the boards are up rather than guess now.” Customers accept this far more readily than trades expect, because it is obviously true.
What all three have in common is that the conversation happens before the work does. What goes on an estimate covers how each of them is actually worded on the page.
One number, and what would change it
Give a single figure wherever you honestly can. A range invites the customer to plan around the bottom of it and hear the top as a failure, so a range only earns its place when the spread is caused by something specific you can name — and then it is usually better expressed as a price plus an allowance.
Whatever the number is, say what would move it. Two sentences is enough, and it converts a later price change from a broken promise into something the customer was told about in advance.
Offer a choice instead of a discount
When a price lands badly, the reflex is to take money off. A better move is to quote a smaller job: the same work with the outbuilding left out, or two rooms now and two in the autumn.
Two or three priced scopes let the customer buy what they can afford without either of you pretending the work is worth less than it is. It also tells you something useful — a customer who takes the middle option was never really price shopping.
The jobs worth walking away from
Estimating is unbilled time, so which jobs you quote is a commercial decision rather than a courtesy. Three signals are worth taking seriously: a customer who wants a ballpark over the phone and will not arrange a visit, a job where you cannot reach the person who decides, and a fifth quote against four you have not seen.
None of those makes someone a bad customer. They just make the odds poor enough that a polite no, quickly, is worth more than a careful quote slowly.
Follow up once
Send the estimate the same day if you can — the gap between the visit and the document is where enthusiasm goes to die. Then follow up once, about a week later, with one short message asking whether they have any questions.
Once. A second and third chase converts almost nobody and costs you the recommendation. If the expiry date on the quote is approaching, that is a natural and non-pushy reason to make contact, which is one more small argument for putting one on.
When they say yes
Get the acceptance in writing, even if writing means a reply to a text. Take the deposit if there is one, and put the date in the diary while you are still in the conversation.
Then the estimate has one job left: to be the thing your invoice is built from. Invoice vs estimate covers how the two documents line up and what changes between them, and how to invoice for services picks the story up once the work is done.
