Hey friends,

A manufacturer told me last month that a good quote takes his team three days. He said it the way you'd mention the weather. Flat, matter of fact, a bit tired. Three days to price a job, apply the margins, work out the freight, put it into a document and send it back to the customer.

I asked him how many quote requests were sitting in the queue behind that one. He went quiet.

That queue is the thing nobody wants to look at. Because while a request sits in it, the customer who asked is off getting three other quotes from people who answered faster.

When people automate their quoting, the first thing they reach for is the cost saving. Look how many hours we got back. That's fine, but it misses the actual prize.

Here is what I care about. If you can turn a quote around in three minutes instead of three days, two things happen. You win a higher share of the jobs you quote, because you got there first and the customer was ready to say yes. And you open the business up to quote far more jobs than you could before, because the desk is no longer clogged.

The three days was never just a cost. It was a ceiling. It quietly decided how much work your business was allowed to take on. Lift it, and the whole thing can breathe.

This is not a robot arm and it is not science fiction. It is a process that watches for a quote request coming into your inbox and then does the work a person would do, in order.

It reads the request. It goes into your product list. If it's a custom job, it reads the drawing or the image and works out what needs to be measured. It checks whether anything has to come from a supplier, and pulls the live supplier pricing. It runs the freight calculation. Then it assembles the quote and puts it in front of a person to check before it goes anywhere.

That last step matters, and I never skip it. A human looks at the quote before it reaches the customer. Always. The point is not to remove your judgement. The point is to hand your salesperson a finished draft to sign off in a minute, instead of a blank page and half a day.

I hear this every time, and it is almost always the reason people never start.

We have built quoting for steel trading, where you have to check the anti-dumping commission, watch the foreign exchange, factor in tariffs and a stack of other moving parts. We have built it for custom manufacturing. We use it on our own proposals, which are custom engineering and consultancy jobs, about as bespoke as it gets.

So here is the test I give people, and it cuts straight through the "too complicated" line. Can you bring in a new person, show them how you quote, and have them doing it properly within a month? If the answer is yes, then AI can do it too. If the answer is no, then the problem was never the AI. Something is wrong with your business. Your product list isn't defined, or your pricing method lives only in one person's head. That is worth knowing either way.

Take the hardest case. Highly customised engineering work, real design thinking, the kind of thing that genuinely needs a human brain. Even there, around eighty percent of the proposal is groundwork that AI can do.

We used to spend two or three days on a proposal like that. Now a tech lead spends about half an hour speaking the solution out loud. This is what I'm thinking, this is how it works, this is what it looks like. The proposal gets built around that expertise.

The result is not subtle. With three tech leads, we used to manage about six proposals a week. We now do around twenty. Same people. Same expertise. We just stopped making them type.

You can have the best marketing team you have ever hired and the sharpest operations in your sector. But if your sales desk is manual, and every quote crawls through it by hand, the business cannot grow past the speed of that desk. Everything you built upstream just backs up behind it.

Most owners have accepted the slow quote as the cost of doing business. It isn't. It's a choice you didn't know you were making.

There is nothing enjoyable about spending three days on a quote that might not even land. Nobody got into their trade dreaming of that. We just decided, a long time ago, that it was simply how things were done, and we stopped questioning it.

The invitation this week is smaller than it sounds. Don't rebuild your whole sales process. Just pick up the last quote your team sent and time how long it really took, start to finish, including the sitting and the waiting. Then ask whether a new starter could learn that process in a month.

If they could, you already have your answer about what's possible. If they couldn't, you've found something more useful than an automation. You've found the part of your business that isn't written down yet.

One thing before I go.

I'm giving a talk called Designing an AI Operating Model at Geelong Design Week, on Sunday 4 October at 12:45pm, at the Nyaal Banyul Geelong Convention and Event Centre.

It's the bigger version of what this whole issue is circling. Everything above, the quote that writes itself, the proposal that gets built around half an hour of a tech lead's thinking, only works when the process underneath it is actually designed rather than just inherited. That's what an operating model is. Not the tools. The way the work is meant to flow, written down clearly enough that a machine or a new hire could pick it up and run it.

If you're in Geelong that Sunday, come along. And come and find me afterwards. I'd always rather have the conversation in person than write another newsletter about it.

Get your tickets here: https://events.humanitix.com/designing-an-ai-operating-model?_gl=1*s2fvij*_ga*MTAxNjM1Mjc5My4xNzg5MDg2OTE3*_ga_2BJRP2W1BV*czE3ODkwODY5MTYkbzEkZzAkdDE3ODkwODY5MTYkajYwJGwwJGgw

See you next week,

— Aamir

🎧 Listen to the Podcast Episode on: Spotify | Apple Podcasts | YouTube

📘 Book: The CEO Who Mocked AI (Until It Made Him Millions) by Aamir Qutub

🚀 Dumb Monkey AI Academy

📱 Dumb Monkey AI Academy App: Apple | Android