How to Follow Up on Estimates With AI: 3 Prompts for Builders and Remodelers

You sent the estimate last week. The homeowner hasn't replied. You're running between jobs, answering questions from the crew, and trying to finish the next bid. Following up is on your list, but it keeps getting pushed to tomorrow.

I've seen how much time owners spend writing the same kinds of emails. AI can take a good chunk of that work off your plate. But there's a catch: if you give it a vague instruction, you'll get a vague, canned email. And that's not how you win a $150,000 remodeling job.

Here's how I'd use AI to follow up on estimates without sounding like a robot or handing over the client relationship.

First, Teach AI How Your Business Communicates

Before you start writing prompts, give your AI assistant some context. Think of it as onboarding a new office coordinator. You wouldn't expect that person to write client emails on day one without knowing your company, your jobs, or how you talk to customers.

Create a short reference file with your company name, the projects you take on, your service area, how you handle estimates, and what makes working with you different. Add a few real emails you've written that sound like you. Tell AI what to avoid, too. Maybe you never say “just checking in,” don't pressure homeowners, and prefer short, friendly notes.

Save that information in your AI project's instructions or reference files. You might call the file “Company and Communication Guide.” Add project-specific notes for active jobs: the client, scope, estimate version, concerns raised, and agreed next step. Keep the facts current.

You don't need a giant manual. A page of useful context beats ten pages of generic brand language.

Three Estimate Follow-Up Prompts Worth Stealing

1. When You Know What You Want to Say, but Don't Want to Write It

This is the easiest place to start. You supply the message. AI helps you get it out the door.

Prompt: “Act as my client communications assistant. Using my company voice guide, draft a short, natural email to Charlie Smith. We sent his kitchen remodel estimate last Thursday. I want to see if he has questions about the scope and offer a 15-minute call next week. Keep it warm and straightforward. Don't sound salesy, and don't add any facts I haven't given you. Give me a subject line, too.”

You can dictate that into your phone between appointments. Read the draft, adjust anything that doesn't sound like you, and send it yourself. No complicated setup required.

2. When AI Already Has the Project History

Here's where it gets more useful. If your AI project has access to the current estimate, meeting notes, and previous correspondence, you shouldn't have to explain the whole job every time.

Prompt: “Review the approved project notes and latest estimate for Charlie Smith's kitchen remodel. Identify the last agreed next step and any unanswered client questions. Draft a follow-up email in my voice that picks up the conversation where we left off. Don't assume Charlie is ready to sign. Don't invent dates, pricing, or availability. If information is missing or contradictory, flag it for me instead of guessing. Show me the key facts you used.”

Notice the difference? You're asking AI to do some thinking before it starts writing. It should know whether Charlie asked for a revised allowance or said he was waiting to talk with his spouse. Those details change the email.

This only works when the information is actually available to the AI assistant. A project folder that isn't connected or uploaded won't magically be read.

3. When You Want AI to Check Your Email and Prepare the Reply

This is the workflow I'd want a busy owner to build toward. Instead of digging through an email thread, you ask the assistant to check it and prepare your next move.

Prompt: “Check my connected Gmail for the latest email thread with Charlie Smith about his kitchen remodel estimate. Tell me whether he has responded since my last message. If he has, summarize what he said and draft a reply in my voice that answers his questions using only confirmed project information. If he hasn't, draft a polite follow-up based on our last exchange. Flag any pricing, scope, or scheduling question that needs my decision. Save a draft for my review if you have permission to do that. Do not send anything.”

For this one, your AI tool needs a working Gmail connection and permission to read the relevant messages. Creating a Gmail draft also requires the appropriate access. If your tool can't do those things, paste the email thread into your project and use the same instructions. You still get a useful draft.

The point isn't to let AI manage your client relationships. It's to spend less time hunting for information and more time making the decisions only you can make.

A Few Rules Before You Hit Send

Never let AI invent a start date, discount, scope item, or promise to a homeowner. Verify the recipient, estimate version, pricing, and any commitments. Review each message before sending, especially when a project is in negotiation. And only share client information through tools and connections your business is comfortable using.

One more thing: the best follow-up is based on the actual conversation, not a generic “touching base” template. If a client asked for time, respect that. If they raised a concern, address it. If they went quiet after receiving the estimate, give them an easy next step.

Try This on Your Next Three Estimates

Start with prompt one. Once you like the drafts, create your company voice guide and move to prompt two. When your email and project information are connected, test prompt three with one client thread. See how much time you save, including review and corrections.

If you'd like help getting your AI assistant set up, start with our AI Toolkit: https://www.thebusinessremodel.com/resources/ai-toolkit/access