How-to

From quote to invoice without typing it twice.

Every page on this site says the same thing about software in a trade shop: the leak is the re-typing between systems, not the systems. This is the page that shows the work. One job traced through five systems with every re-type counted, then the three fixes in the order they should be tried, and the arithmetic for when each is worth it.

No email required. No gate. The trace is an illustration with round numbers, not a client and not a statistic. Your count will differ, which is the point of counting.

01

The trace: one job through five systems

A water-heater replacement at a plumbing shop, because every trade has a version of it: a short, priced job with a part in it. The shop owns a field-service tool and QuickBooks. It also owns a phone pad and a quote template, which is where the job actually lives. Read the right-hand column.

  1. 01 · The phone

    The call comes in

    The office writes the name, the address, the phone number and “no hot water, 50-gallon gas” on the pad, or in the notes field of whatever the phone system has.

    Typed for the first time

    Name, Address, Phone, The problem

    Typed again

    Nothing yet

  2. 02 · The quote template

    The call becomes a quote

    The office opens the Word or Excel template, types the name, address and phone again, picks the heater and the parts from a supplier PDF, types the prices, and emails a PDF.

    Typed for the first time

    Line items, Prices, Tax

    Typed again

    4 facts

  3. 03 · The field-service tool

    The quote becomes a job

    The customer texts “yes.” The office creates the job: name, address, phone again, “WH swap 50 gal,” a tech and a slot. The line items do not come across; the job note says “see quote.”

    Typed for the first time

    Tech, Time slot

    Typed again

    4 facts

  4. 04 · The invoice

    The job becomes an invoice

    The tech texts “done, used the 50, plus an expansion tank, 3.5 hours.” The office builds the invoice from the text and the quote PDF: the lines again, the prices again, the tax again, and if the invoice lives in the books rather than the tool, the name and address again.

    Typed for the first time

    The extra part, Hours

    Typed again

    7 facts

  5. 05 · The books

    The invoice lands in the books

    The card is taken over the phone in the tool. At month end the bookkeeper enters the invoice and the payment into QuickBooks by hand from a report, because the sync was never switched on, or was switched on once and turned off after it did something unexpected.

    Typed for the first time

    Nothing

    Typed again

    9 facts

Five systems, four arrows, and the job’s facts were typed by a person for the first time exactly once each. Everything in the “typed again” column, 24 re-types on one job, is somebody’s normal Tuesday.

02

The count

Put round numbers on it. Re-typing a name, an address and a phone number into a new screen is a couple of minutes if nothing goes wrong. Re-building a quote’s line items from a PDF is longer, call it eight, because the prices have to be found again and the tax has to be right. On the trace above that is something like forty minutes of pure re-typing per job, before a single error.

Now the errors, which are the expensive part and the part nobody counts. The address typed four times will be typed wrong once, and the tech will go to the wrong house or the invoice will go to the wrong email. The line items rebuilt three times will drift: the expansion tank on the invoice and not on the quote, the tax on one and not the other, the price from last year’s PDF. The month-end entry will not match the tool, and the bookkeeper will spend a morning finding out why.

Forty minutes a job, at twenty-five jobs a week, is around seventeen hours a week of somebody in the office typing what somebody else already typed. That is a part-time salary. It is not a statistic; it is the multiplication for the trace above, and yours will be different. Do the multiplication for one of your own jobs before you read the fixes, because the fixes are ranked by which arrow costs the most, and that is a number only your shop has.

The calculator does the multiplication from four inputs →

03

The three fixes, in order

The order matters more than the fixes. Every vendor will sell you the third one first, because it is the one with a price on it. Most shops never need it.

  1. 01

    Configure what you own

    First, always. Most of the re-types in the trace are inside one product’s reach.

    Build the quote in the field-service tool instead of the template, which means loading the price book, the afternoon nobody has had. Then the accepted quote converts to the job with the lines on it, the closed job converts to the invoice with the lines on it, and the QuickBooks sync, switched on and left on, carries the invoice and the payment to the books. In the trace that removes three of the four arrows: the contact details are typed once, at the phone, and the line items once, at the quote.

    What it costs
    A price-book afternoon, a settings hour, and a rule for the bookkeeper: edit in the tool, read in the books.
    What it removes, on the trace
    Arrows two, three and four, and most of five.
  2. 02

    Bridge what you cannot

    When a step has to live outside the tool, and configure cannot reach it.

    The quote comes from a takeoff or estimating product, or a supplier’s quoting portal, or a spreadsheet the owner will not give up because it does the option pricing the tool cannot. Fine. Then the accepted quote is pushed into the field-service tool as a job with its lines by a connector: the vendor’s native integration if one exists, an off-the-shelf automation tool if the mapping is field-to-field, and a small custom bridge when the mapping needs logic, a kit that becomes six line items, an option set where only the chosen one crosses. That last one is what “I build inside the tool you already run” means in practice.

    What it costs
    Hours for a native integration or a simple map; days for a bridge with logic. Priced as a fixed-scope build when it is more than a map.
    What it removes, on the trace
    The one arrow configure could not: the quote into the job.
  3. 03

    Build only what nothing can hold

    Last, and only when a step cannot be represented in any tool in the chain.

    The quote is assembled from three photos, a voice note and the price book, and no tool does that. The job type needs a form or a calculation the field-service product cannot hold. The contract has to be read before the quote is written. These are the builds, and they are small, scoped, and inside the tools you already run, not a new system beside them.

    What it costs
    Fixed scope, $5,000 to $10,000, after the audit has shown the arrow is worth it.
    What it removes, on the trace
    A step that was manual because nothing could do it, not because nothing was configured.

Run the trace again after the first fix and the water heater looks like this: the call is typed once, the quote is built once from the loaded price book, the job and the invoice inherit it, and the books receive it. Two of the five systems are gone, the pad and the template, and nobody misses them. That is the outcome most shops are actually buying when they buy AI, and it did not need any.

04

When each one is worth it

One piece of arithmetic, done per arrow, not per shop. Re-types a week on that arrow, times minutes each, times the loaded cost of the person doing them. That is the leak. Set it against the fix.

Under an hour a week: configure it if a setting reaches it, and otherwise leave it alone. An arrow that costs an hour a week is not worth a bridge that costs days to build and a Tuesday every quarter when the vendor changes a field name.

A morning a week: configure first, always. If configure cannot reach it because the step genuinely lives outside the tool, bridge, and expect the bridge to pay for itself inside a quarter.

A day a week or more on one arrow: this is the arrow that justifies a build, and it is usually the quote, because the quote is where the most facts are created and the most judgement is applied. Build it only after configure and bridge have been tried on the same arrow and have fallen short for a reason you can name, not because the demo was good.

The trap in the other direction: a shop that will not do the price-book afternoon and buys a bridge instead. The bridge faithfully pushes a quote with no lines on it into a job with no lines on it. Configure is not optional because it is cheap; it is first because everything after it depends on it.

05

What the audit adds

Everything above you can do with a notepad and a Tuesday. What you cannot easily do from inside the shop is do it for every job type, with the people who do the typing rather than the person who thinks they know, and with a dollar figure beside each arrow that the owner will believe. That is the AI Audit: the trace, done across the shop in interviews, the count with real minutes and real loaded costs, and a written blueprint that says for each arrow configure, bridge, or build, ranked by what it recovers. It also says which arrows to leave alone, which is the part no vendor will write.

A fixed fee, $5,000, sized by headcount. If the recoverable savings in the blueprint do not cover it, you do not pay, and you keep the blueprint. The first step is a 30-minute call, and I will say on the call if the trace above is all you need.

06

Questions people actually ask

What does “double entry” mean in a trade business?

Not the accounting kind. It is the same fact, a customer’s address, a line item, a price, typed by a person into a second system that could have received it from the first. In most shops a job passes through four or five systems and each arrow between them is a re-type. The fix is not one big system; it is making each arrow carry what the previous step already knew.

How do I stop typing the same job into three systems?

Trace one job and count where it happens, then take the fixes in order. Configure: build the quote inside the field-service tool so it converts to the job and the invoice with the lines on it, and switch the QuickBooks sync on. Bridge: if the quote has to live elsewhere, connect it so the accepted quote arrives in the tool as a job. Build: only for a step no tool can hold. Most shops stop after the first.

Does Jobber or Housecall Pro convert a quote to an invoice?

Both do, and so does every field-service product in the category: an accepted quote becomes a job with the line items carried across, and a closed job becomes an invoice the same way. The re-type happens when the quote was built outside the tool, in a template, because the price book was never loaded. Neither product can fix that, and it is the first thing to fix.

Should I use Zapier or Make to connect my tools?

For a field-to-field map, yes: a form submission becomes a lead, a paid invoice posts a message. For a quote becoming a job, usually not on its own, because the mapping has logic in it: which option the customer chose, which kit expands to which lines, what happens when the customer edits the quote after accepting. When the map needs an if, it needs a small bridge with the if written down, and someone who will own it when the vendor changes a field.

When is it worth building something custom?

When the arrow is expensive and nothing configurable or bridgeable can carry it. Count first: re-types a week, minutes each, the loaded cost of the person doing them. If the leak is under an hour a week, configure and stop. If it is a day a week and configure cannot reach it, bridge. Build when a bridge would be pretending a tool can hold a thing it cannot, and never before the count.

How much does a bridge cost?

A native integration is included in the tool you pay for. A simple map in an automation tool is a monthly fee and an afternoon. A custom bridge with logic, or a build for a step nothing holds, is fixed-scope, $5,000 to $10,000, inside your tools. Finding out which arrow in your shop is worth which fix, with a dollar figure beside each, is the AI Audit: a fixed fee, $5,000, sized by headcount, not owed if the savings do not cover it.