← The Maker Growth Blog

A custom order DM just landed. Here's how to turn the mess into a quote.

AI for makersChatGPTcustom ordershandmade business

“Hi! Do you do custom ones? I need something for my sister 😊” That’s the whole message. No quantity, no date, no idea what “something” is. If you make anything personalized, you know how custom order requests actually arrive: half a thought in an Instagram DM, an Etsy message, or a Facebook comment, and it’s on you to turn it into an accurate quote without a twelve-message back-and-forth. Learning how to handle custom order requests isn’t about writing faster replies. It’s about having a system that catches the missing details before they cost you a bad quote or a remake. Here’s the one that turns a messy inquiry into a clean brief, a professional reply, and a checklist you can actually build from.

That message is made up, but it isn’t exaggerated. Most custom inquiries look like it, wherever they arrive.

Why custom requests go sideways: the missing half

Almost every custom-order headache traces back to one thing: you quoted or started before you had all the facts. The buyer knows what they want in their head; they just didn’t type it. Quantity, exact dimensions, colors, the design file, personalization text, packaging, where it’s shipping, the in-hand date, the budget — any one of those, left unspoken, is a remake or a refund waiting to happen. So the first job isn’t to reply fast. It’s to separate what you actually know from what you’re still missing, and only then respond. That’s a sorting job, and sorting is exactly what AI is good at. It’s the same reason it’s one of the first AI workflows every maker should set up.

Step 1: redact, then paste the inquiry into ChatGPT

Before the message goes anywhere near a chatbot, strip the personal stuff — names, addresses, phone numbers, anything payment-related. You’re pasting the request, not the customer’s identity. This is a small habit that keeps you on the right side of your buyers’ privacy, and it costs you ten seconds.

Step 2: run the intake prompt, confirmed vs. missing

Now put AI to work as an order-intake assistant with a prompt that forces it to sort, not invent:

Act as an order-intake assistant for a custom maker business. From this inquiry, separate confirmed details from missing details. Check for quantity, dimensions, material, design file, colors, personalization, packaging, delivery location, in-hand date, budget, and proof approval. Do not invent answers or assume anything I didn’t say.

The “do not invent” line matters. Without it, AI will cheerfully fill gaps with plausible guesses, and plausible guesses are exactly what burn you on a custom order. What you want back is two clean lists: here’s what the buyer told you, and here’s the hole in the brief. That grounded, spell-it-out approach is the whole point of prompting ChatGPT so it stops handing you confident wrong answers.

Then add the details your own craft can’t quote without. A laser shop needs the exact blank and its size. A UV shop needs to know whether the surface is curved and whether the design needs white ink. Anyone working from customer artwork needs the file itself, at a resolution good enough to use. That last one is where a lot of free labor hides. In my own shop, file cleanup was the step I kept giving away, until I moved uploads into my Shopify store with PartyKit Uploads. Now the upload is part of the order, and the file is checked for resolution and quality before it reaches me.

Step 3: draft the warm reply with only the questions you need

Here’s where most makers overcorrect and fire back ten questions, which reads like a form and kills the excitement. Instead, ask AI for the minimum:

Draft a warm, friendly reply that answers their enthusiasm and asks only the fewest questions needed to prepare an accurate quote. Keep it short and human.

You want a reply that sounds like a maker who’s glad they reached out — not an intake bot. Two or three well-chosen questions (“Lovely idea! Two quick things so I can quote it properly: how many, and when do you need it in hand?”) move the order forward without overwhelming them. Edit it so it sounds like you; AI gets you 80% of the way in seconds, and your voice does the rest.

Step 4: build the quote-prep table and the proof checklist

Once the buyer fills the gaps, have AI organize — never decide — the rest:

Now produce a quote-preparation table (line items I should price) and an internal production checklist for this order. Do not calculate prices or promise a delivery date.

The table is a worksheet, not a quote: it lists what to price so you don’t forget the rush fee or the second proof. You plug in your own documented rates. The checklist is your production safety net — file received, proof sent, proof approved, materials on hand, packaging, ship-by date. The same structured-reply-plus-follow-up discipline scales straight up to bigger jobs, which is exactly how to quote corporate gift orders without dropping a detail.

Two checklist lines are worth stealing. First, never cut or print before the proof is approved in writing, and that includes the spelling of every name. Second, check the file before it touches a blank. The free Laser Cut File Checker catches open paths, doubled lines and text that was never converted to outlines, and the UV DTF Print Checker does the same job for UV artwork. For a product that’s new to you, add one more line: run a couple of test pieces before the customer’s piece goes on the bed. I do that with every new product in my own shop.

The line AI never crosses

AI can sort the inquiry, draft the reply, and lay out the checklist. What it must never do is set the final price, promise a delivery date, or decide a file is production-ready. Those three are judgment calls that depend on your rates, your queue, and your machine — get any of them wrong and you’ve either lost money or lost a customer. Pricing comes from your documented numbers. Dates come from your real capacity this week. File approval comes from your eyes on the artwork. Let AI handle the organizing so you have the head-space for the deciding. That’s the healthy split for your whole back office, not just custom orders.

Here’s how a missing detail usually plays out (an example, not a real customer): a buyer orders a cutting board engraved “for my sister” and approves the proof. Nobody asked about the in-hand date, and it turns out the gift is for a party in three days. Now it’s a rush job at the regular price, and the shipping upgrade comes out of your margin. One question in the first reply, “When do you need it in hand?”, would have priced the rush in from the start.

Sort it with AI. Quote it yourself.

Keep reading

Ready to put this to work? Join the Maker Growth Hub.

Build this into your shop

The Maker Growth Hub gives you the templates, prompts, live calls, and community to put tactics like this to work — $25/mo, locked for life.

Join the Hub →
← All posts