A system that supports quoting

Why quotes sit in a queue

Monday, 7:40. An enquiry from a new client is waiting in the inbox: a PDF with a table of line items, two drawings and a request for a quote by Friday. The person who knows how to price it is tied up with current work right now, because something has needed their attention since the morning. They will get to the enquiry on Wednesday evening. On Thursday the sales rep sends the quote, and on Friday he hears that the client has already agreed terms with the company that replied on Tuesday.

This scenario repeats itself in companies regardless of industry. And it has surprisingly little to do with price. The order is lost on response time, before the conversation about money even begins.

Quoting is expert work. Someone has to read the specification or drawing, assess the material and components, estimate times and labour, and check what similar work the company has already done. That knowledge usually sits with two or three people - the process engineer, the department manager, sometimes the owner themselves. They are also the people most tied up with day-to-day work, so the enquiry lands at the bottom of their to-do list.

On top of that, the data needed for a quote lives in several places at once. Old quotes in Excel files, actual lead times in the system or on paper job cards, material prices in emails from suppliers. Every quote therefore starts with searching and only then with calculating. The result is predictable: replies go out after several days, some enquiries never get a quote at all, and the company does not even know how many orders a year it has lost to the calendar alone.

A solution in three parts

Exactly this process can be built today into a company's system with the help of AI. The solution has three parts.

First, reading the enquiry. With the help of AI, the system reads the message and its attachments: the table of line items, quantities, material, dimensions, tolerances, the expected deadline. The email becomes a record in the system, complete before anyone sits down at a keyboard. Manual retyping disappears, and with it the typos that later follow the order all the way through. When something is missing from the enquiry, the system proposes a list of follow-up questions for the client and immediately prepares an email with those questions, a document to fill in or an interactive web form that only needs to be shared with the client. The answers come back into the system and go through the same AI analysis as the enquiry itself, so the quote is a better fit for what the client actually needs.

Second, matching against history. The system searches completed orders and earlier quotes: a similar part, a similar material, a comparable batch size. It brings up the actual times and costs from delivery, and next to them the price and margin from that earlier quote. A reference point like this exists in every company with an order history, only today it has to be dug out of files and out of people's memory.

Third, a draft quote for a human decision. From the enquiry it has read and the history it has matched, the system assembles a draft quote: line items, proposed times, material cost, a space for the price. The decision on price and on risk stays with a human - the process engineer, the sales rep or the owner, wherever that is required. The machine reads and calculates, the human prices. The system also records which quotes were won and which were lost, so after a few months the hit rate of the quoting is visible in black and white.

What this means for the business owner

The calculation is best done on your own numbers. How many enquiries come into the company each month? How many hours of work does one quote cost, and what are those hours worth, given that the person doing the quoting could be doing something else for the company in that time? Often it is the owner themselves who quotes, and their hours are the most expensive in the company. After how many days does the client get an answer?

And the most important question: how many orders a year are lost because the quote was sent too late? It is enough to take the value of one average order and work out what one extra order won per month changes. With orders that come back on a regular cycle, that calculation covers whole years of working with a client who once had to choose the faster supplier.

There is a second side to this calculation too: a company that responds quickly and on the basis of actual costs from its history quotes by gut feel less often. And a gut-feel quote can cost more than the most expensive system, because one time it sets the price too high and scares the client off, and the next time too low, and the company subsidises the order.

Where to start

The first step is simple and needs no AI: a register of enquiries and quotes in one place, with the date received, the date answered and the outcome. That register alone shows after a month where the time goes and how many quotes actually leave the company.

The second step is a repeatable quote structure: line item, material, times, costs, margin, stored in the system instead of in loose files. Only on data that orderly is it worth adding automatic reading of enquiries and suggestions from history. Automation makes sense on orderly data; layered on a mess, it only makes the mess bigger.

There is one more benefit that rarely comes up when people think about quotes: the look. A quote produced from the system can simply be beautiful - in the company's visual identity, with the logo, a clear table of line items and a summary, the same every time. The AI generates the document from an agreed template, so the quality of the quote stops depending on how well someone in the company gets on with Word.

If you would like to see what a quoting workflow like this could look like in your company, write to me or book 30 minutes straight away in the calendar below.

More articles about what can be built in your company are in the Knowledge section.