6 min read
This is an illustrative composite story, based on common workflows and real, currently available tools. Marco is not a real person, but every step he takes is something you can try this week.
It is 7:40 on a Tuesday evening, and Marco is sitting in his truck outside his fourth estimate of the day. His phone shows 11 unread texts, two voicemails, and a camera roll full of overgrown hedges and patchy lawns. Somewhere in that pile are three people ready to hire him, if he can get them a price before the landscaper down the road does. He knows from experience that he won’t. Not tonight. The quotes will go out Thursday, maybe Friday, and at least one of those customers will have already said yes to someone else.
The Real Problem Was Never the Mowing
Marco runs a one-person landscaping business: mowing, hedge trimming, spring cleanups, and the occasional small planting job. He is good at the work and his regulars love him. His problem was the paperwork that wraps around the work.
Every new job followed the same slow loop:
- A customer texts photos and a vague request (“can you make the backyard not look like a jungle?”).
- Marco drives over to look, because the photos never show enough.
- He writes rough notes, then types a quote at night when he is exhausted.
- He forgets to follow up on half of them.
- When the job is done, he invoices days later.
He estimated he was losing about one job in three simply because his quotes arrived late. Sound familiar? It is the same pattern we saw in the story of a solo real estate agent who kept losing leads to slow follow-up. Different industry, same leak.
Step One: Teaching the AI His Prices
Marco didn’t start with anything fancy. He opened ChatGPT and spent one evening writing down how he actually prices things: his rate for mowing by approximate lawn size, hedge trimming by length and height, cleanup by number of yard waste bags, his minimum charge, and his travel surcharge for addresses outside town.
He saved all of it as a short price sheet and pasted it into a reusable instruction, so every new conversation started from the same rules. This was the single most valuable hour of the whole project. An AI cannot quote your work until you can explain how you quote your work.
Step Two: Photos In, Draft Quote Out
Now when a customer texts photos, Marco forwards them into ChatGPT with a quick note: “Front and back lawn, hedges along the left fence about 30 feet, customer wants a one-time cleanup plus biweekly mowing.”
The assistant looks at the photos, estimates the scope against his price sheet, and drafts a quote with line items, plus a friendly message to the customer. Then comes the part Marco insists on: he checks every number before anything goes out. Photos can be misleading. A hedge might be taller than it looks, or a slope might make mowing slower. About a third of the time he adjusts something.
The difference is when it happens. Instead of typing quotes at 10 p.m., he reviews AI drafts between jobs, sitting in the truck, in about five minutes each. For bigger jobs he still visits in person, but now he arrives with a draft already in hand.
If you want to build something similar for a different kind of service, our guide on building an AI proposal workflow that sends in minutes walks through the general version.
Step Three: A Simple Lead Tracker That Updates Itself
Marco’s second leak was follow-up. He had no system, just a mental list that got shorter every time he got tired.
He set up a basic Google Sheet with columns for name, address, job type, quote amount, date sent, and status. Every time he finalizes a quote, he asks the assistant to give him a one-line summary in the same format and pastes it in. Google has been expanding how its Gemini assistant connects to other tools (its September update added connections to apps like Airtable and monday.com), so this kind of tracking is getting easier to automate further, but Marco kept it manual on purpose. Simple systems survive busy weeks.
Every Monday morning, he pastes the sheet into ChatGPT and asks: “Which quotes are older than five days with no answer? Draft a short, friendly check-in text for each.” He edits, sends, done. For more on sorting incoming requests automatically, see how to build an AI intake flow that sorts your leads.
Step Four: Invoicing Before He Leaves the Driveway
The last piece fell into place this fall. Marco already used QuickBooks for his books, and Intuit’s September updates added the ability to create and send QuickBooks invoices and payment links directly from ChatGPT or Claude.
Now, when he finishes a job, he types one line into the same chat where he drafted the quote: “Job done. Invoice the Hendersons for the cleanup as quoted, plus two extra yard waste bags.” He reviews the invoice, sends it with a payment link, and the customer often pays before Marco reaches the next address. Since some of these AI features are tied to specific QuickBooks plans and are still in beta, he checked his plan first.
What Changed After Two Months
In this illustrative scenario, the results look like what many trades owners report when they tighten the quoting loop:
- Quotes go out the same day for most jobs instead of two or three days later.
- Fewer lost jobs. Faster quotes and steady Monday follow-ups mean far fewer customers drift to a competitor.
- Evenings back. The late-night typing session is gone. Most admin now happens in short gaps between jobs.
- Faster payment. Invoices sent on site get paid sooner, which matters for a business that buys fuel and equipment upfront.
Just as important is what did not change. Marco still walks every big property. He still sets his own prices. He still calls his regulars personally. The AI handles the typing, not the relationships.
What Marco Would Tell Another Tradesperson
- Write your price sheet first. Everything else depends on it.
- Never send an AI quote unread. Photos lie. Your experience doesn’t.
- Keep the tracker boring. A spreadsheet you actually use beats an app you abandon.
- Follow up on a schedule, not a mood. One fixed time per week is enough.
- Invoice before you leave. The best time to get paid is when the customer is looking at a freshly finished yard.
None of these steps requires technical skills, and none costs more than the tools a small trades business probably already pays for. If you run a service business that lives on quotes, which one of these steps would save you the most time this week?
