Most estimates are not turned down. They are just sitting there.

I asked a builder what his conversion rate on estimates was. Four in ten, he said, straight away, no hesitation. Good number, he was pleased with it.

So we went and counted. Out of a hundred estimates over the previous year: forty one won, which was fair enough, he knew his business. Eleven had come back as a no, mostly on price. And forty eight had never been answered at all. Not declined. Just nothing, ever, in either direction.

He had been treating those forty eight as losses. In his head they were people who had looked at his price and gone elsewhere, and he had spent a fair amount of time over the years quietly wondering whether he was too expensive. When we actually rang a dozen of them, the answers were: I forgot, I never got it, I was waiting for the wife to decide and then it just drifted, we did it ourselves in the end, and one man who could not remember the conversation at all.

Not one of them said the price was too high.

Silence is not a no, it is an unfinished job

This is the single most expensive misreading in trades and service work. Silence gets filed as rejection because rejection is a closed loop and closed loops are comfortable. If they said no, there is nothing to do and nothing to feel awkward about.

But an unanswered estimate is a job that stopped halfway through, and it stopped for reasons that are almost never about you. People are busy. People are waiting on a partner, or a bonus, or for the weather. People genuinely do not see the email, which is worth sitting with for a second, because the estimate that took you an hour to price and forty minutes to drive to is, on their phone, one grey line in a list of forty.

Same day, or you are competing with a memory

The estimate should leave the same day as the visit. Not because of some rule about being professional, but because of what the customer is doing in the meantime.

On the day, you are a specific person who came to their house and knew what they were talking about and pointed out the thing under the sink nobody else spotted. That is a real advantage and it is worth actual money. Three days later, you are a number in an email from a company they cannot picture. Whatever goodwill you built by turning up on time and being good at your job has fully evaporated, and you are back to being compared on price alone, which is the one comparison you least want.

Most people know this. The reason estimates take four days is not that anyone thinks four days is a good idea, it is that pricing happens in the evening after a full day of work, and the evening is where things go to die. So the fix is not discipline, it is making the estimate mostly assemble itself: the measurements and photos captured on site, the line items and rates already known, so the evening job is checking a number rather than building a document from a blank page.

The follow up nobody sends

Here is what most businesses do after sending an estimate: nothing. Here is what a few do: one awkward phone call about a week later, if they remember, which is the most uncomfortable phone call in the trade and gets avoided accordingly.

The version that works is three light touches and then a full stop.

1. 01Two days later. Short, no pressure, purely practical. Did that come through alright, and is there anything in it you want me to explain. The job of this message is to catch the ones who never saw it, and it catches a surprising number.
2. 02About a week later. This one adds something rather than repeating the ask. A note about timing is usually the most useful thing you have: I have got space week after next, after that it gets tight. True scarcity, if it is true. Never invented scarcity, because people can smell it.
3. 03Two or three weeks in. The closing message, and the one people are most nervous about sending. I will assume this one is not going ahead so I stop cluttering your inbox, but if it comes back around just reply to this and I still have the figures. Then stop.

That third message is the one that gets replies, consistently, and it is the one everybody wants to cut. Giving someone permission to say no gets you an answer, and an answer is what you actually needed. A clean no lets you stop wondering, and it lets you plan the month. About a fifth of the time, in my experience, the reply is not no at all, it is sorry, yes, when can you start.

Then the other awkward chase

The second half of this is the invoice, and it has exactly the same disease. The work is done, the customer is happy, the invoice is out, and it is now three weeks old, and asking about it feels like accusing a nice person of something.

So most people wait, and then wait a bit longer, and then send something apologetic with the word just in it three times. Meanwhile the customer has not paid because it slipped down the list, or because they are waiting for the invoice to remind them, which it will not, because invoices do not do that on their own.

Automating the reminders removes the emotion from it, which is the actual point. Nobody is annoyed by a scheduled reminder about an invoice, because everybody understands that a business sends those. It is not a personal accusation from someone they had a nice chat with about their kitchen. Reminders keep going, politely, on a schedule, until it is settled, and you do not have to feel anything about any of them.

One practical thing while we are here: make paying it take one tap. A bank transfer that requires typing an account number, a sort code, and a reference is three chances to give up and do it later.

What we deliberately do not automate

The price. Not because we could not build something that spits out a figure, but because the figure is your judgement and your risk, and the moment a system is choosing it you have handed over the one decision that decides whether the job is worth doing.

What gets automated is everything around it. The document, the sending, the chasing, the conversion into an invoice, the reminders. You still decide what the job costs, on site, with the information in front of you. Anyone offering to take that off your hands is selling you something you should not buy.

Where this does not work

• Long procurement. If your customer is a council or a main contractor with a tender process, none of this applies. Their timeline is not affected by your follow up sequence and a nudge at day two will irritate somebody.
• Genuinely complex quotes. If pricing a job requires three site visits and a structural opinion, the constraint is not your admin and speeding up the paperwork changes nothing.
• Very high value, very rare work. Two estimates a month should be followed up personally, by you, on the phone. A sequence is for volume, and at that volume you have the time.
• When the real issue is the price. If your win rate on answered estimates is low and people are telling you why, follow up will not fix it, it will just get you clearer nos faster. Which is information, but it is not the information you were hoping for.

This sits on top of records that look after themselves, which is why we usually do the spreadsheet work first or at the same time. An estimate that chases itself needs to know when the job was completed, and that has to come from somewhere.