Winback for Builders
The customer cannot come back. The architect who stopped calling can
Winback means something different in your trade
A single storey extension is £19,450. A dormer loft conversion, £17,200. A garden room, £12,850. A garage conversion, £5,900. An internal wall removal with an RSJ, £2,750.
In most trades a winback campaign goes to lapsed customers. For a builder that is largely a waste of an afternoon. The person who spent £19,450 extending their house in 2022 is not going to do it again, and no amount of well-written contact will change that.
Your lapsed population is somewhere else entirely, and it is made of professionals.
The architect who specified you on two jobs in 2023 and none since. The structural engineer who used to recommend you. The estate agent who sent people your way for a year. The kitchen firm that stopped mentioning your name. Each of those relationships was worth a great deal, each of them stopped, and in almost every case nobody ever found out why.
That is the recoverable money in a builder’s business, and it is invisible because nobody fired you. The work simply stopped arriving and you were too busy at the time to notice.
Why they actually stopped, which is rarely what you would guess
In our experience it is almost never price and almost never quality.
Somebody else turned up. A newer firm rang them, took them for a coffee, made themselves easy. Yours was a relationship nobody maintained.
You were unavailable once. They rang about a job, you were eighteen weeks out, they used somebody else, and that firm has had the next six.
One job went badly at the end. A snagging list that dragged, a variation argument, a final account that took four months. The building was fine; the finish of the relationship was not.
The person moved. Architects and agents change practice constantly. Your contact went somewhere else and nobody replaced them in your address book — and, worth noting, they are now recommending builders in a new firm.
Only one of those is about your work. Three are about attention, which is exactly what a winback programme supplies.
The ratio
An extension is £19,450. Our Booked package is £99 a month — £1,188 a year. One extension covers more than sixteen years.
The winback arithmetic is the strongest on this site. A lapsed architect who used to specify two jobs a year is worth roughly £38,900 a year at your average value. Recovering one of them justifies every other thing we could sell you, several times over, and the work involved is a list, four conversations and a follow-up routine.
What we’d actually do
Build the lapsed list properly. Every professional and referrer who has ever sent you work, with the date of the last job they were involved in. Most builders have never assembled this and are startled by how long the list is.
Find out what happened, by asking. A short, direct, unembarrassed conversation — we noticed we have not worked together since the Mill Lane job, is there something we could have done better. Builders hate this call. It is the highest-value twenty minutes in the whole programme and the answers are usually mundane and fixable.
A reason to reappear that is not a sales call. A finished job worth showing, a regulation change that affects their advice, a slot in the programme. Something useful gives the contact a reason to reply.
Chase the person, not the practice. When your contact has moved firms, that is two relationships rather than a lost one, and almost nobody follows them.
A standing rhythm afterwards, so the recovered relationship does not lapse again in eighteen months for exactly the same reason.
What good looks like at twelve months
- A complete list of referrers with the date they last sent you anything
- Four to six honest conversations with people who went quiet
- At least one relationship visibly restarted
- Contacts who changed practice tracked rather than lost
- A quarterly rhythm that stops the next lapse happening silently
The mistakes we see most
Assuming they were unhappy. Most of the time nobody was unhappy. You were busy and somebody else was present.
Never asking why. The question feels awkward for about four seconds and then produces an answer you can act on.
Only chasing when work is short. It is transparent, and it is why the call lands badly in February and well in a busy month.
Letting a contact’s move end the relationship. They have taken their specifying habit somewhere new and nobody there knows you.
The first ninety days
Weeks 1–4. We assemble the referrer list with the last-job dates, which is where the surprises are.
Weeks 5–8. The reappearance material — a project note worth sending — and the first round of conversations with the people who have gone quietest.
Weeks 9–12. Moved contacts traced, and a quarterly rhythm set so this becomes maintenance rather than rescue.
At ninety days the measure is whether any lapsed referrer has put your name forward again. In a trade where one relationship carries £38,900 a year, one is a good quarter.
Where this lands in Gaffer
Contacts sit under the practice they work for, so a referrer who moves firms remains a person you hold rather than a name attached to a company that has stopped calling.
Jobs record who introduced them, which is what makes a lapsed-referrer list a query rather than an act of memory — you can see exactly who has sent nothing for two years.
And because quotes hold their outcome, the conversation about what happened can start from what actually happened rather than from your recollection of it.
The streets you have already worked
There is a second dormant population, and it is not made of people who have ever paid you anything.
Every job you have run put you on a residential road for three to six months. A few hundred households watched the whole thing — the steel going in, the roof going on, the skip changing over — and formed a view about the firm doing it. Several of them have been thinking about their own extension ever since, because seeing one built next door is the most common trigger there is.
Those streets are dormant in exactly the way a lapsed customer is. There was contact, there was interest, and then you finished and left and nobody ever went back.
Going back is straightforward and nobody does it. A short letter to thirty or forty houses on a road you worked eighteen months ago: we built the rear extension at number 12 in 2024, we are working in the area again this spring, here is roughly what that kind of job involves and costs. It refers to something they watched happen, which is why it gets read.
The houses are similar, the constraints are similar, and you already know the local planning position and what the ground is like. That last part is worth stating in the letter, because it is the single most credible thing you can say to somebody on that road.
Questions builders ask
Is it not obvious I only rang because I am quiet? Only if you ring when you are quiet. Run it as a rhythm and the question never arises.
What if the answer is that a job went badly? Then you have learned something worth a great deal, and an acknowledgement is frequently enough to restart a relationship.
Should I bother with past homeowners at all? As referrers, yes. As repeat buyers, very rarely — which is the whole argument of this page.
How many should I contact? All of them, over a quarter. It is a short list and every name on it can send you a £19,450 job.
Get your free website review · Bigger outfit? Talk to us