How to use AI in a solar installation business in 2026

Solar is a paperwork business wearing a hard hat. A residential job runs through qualification, a site visit, a design, a proposal, a permit, an install, an inspection and a utility interconnection before the homeowner sees any benefit. Every handoff creates questions, and most land on your office as calls and texts.

That makes solar a good fit for AI in one specific way: drafting the words that move a job from one stage to the next. It also makes solar a risky fit in another way, because the numbers homeowners care about most, production, savings and incentives, are exactly the numbers a chatbot will confidently get wrong. This guide shows where the line sits for a $2M to $30M residential solar installer. If you want the broader picture first, read AI for contractors in 2026: what actually works.

What should AI do in a solar business, and what should it never do?

AI writes first drafts of words. People own designs, numbers and promises. Here is the rule sheet we would hang in a solar office.

How can AI help qualify solar leads?

Solar leads are expensive and many are not a fit. A shaded lot, a roof that needs replacing first, or a renter costs a consultant an evening. AI helps here as a script writer, not as the one asking.

Give a general assistant your real disqualifiers and ask it to draft a short, friendly set of questions your setter can ask in under five minutes. A starting set looks like this:

• Ownership: Do you own the home, and are you the person who makes decisions about it?
• Roof age and condition: Roughly how old is the roof, and has it had any leaks or repairs? A roof near the end of its life often needs replacing before panels go on.
• Shade: Are there large trees or buildings that shade the roof for much of the day?
• Usage: Would you be comfortable sharing a recent electric bill so the designer can work from real usage?
• Payment path: Are you thinking about paying cash or using financing from a lender you choose?

Put the approved questions into your lead form and your intake notes so every setter asks the same things. In Revcore, forms and lead capture feed straight into the CRM and pipeline, so answers sit on the lead record when the consultant opens it. When speed matters, pair this with the practices in our speed-to-lead guide, which also covers texting consent.

How do you use AI for homeowner education without making claims?

Homeowners arrive with half-remembered ads and questions about net metering, batteries and incentives. A written explainer they read before the appointment saves your consultant time on basics. AI is good at turning your knowledge into that explainer and bad at facts about your utility, your state or the tax code.

Incentives are the clearest example of why. The IRS states that the residential clean energy credit is not available for property placed in service after December 31, 2025, per its residential clean energy credit page. A chatbot trained on older material may still describe that credit as if it were available. A homeowner who signs because of a sentence like that has a real complaint, and you own it.

• Write education around process, not outcomes. How a site survey works, what a permit is, what an inspection checks, why interconnection takes time.
• Leave numbers to the proposal. An explainer can say the designer will estimate production for this roof. It should not quote a typical saving.
• Route incentive questions to a person. Someone on your team checks current federal, state and utility programs on official sites, and the homeowner is told to confirm tax questions with their own tax advisor.
• Date it and keep one version. Put a reviewed-on date on the approved explainer, save it in your templates, and recheck it when rules change.

How can AI help prepare a solar proposal presentation?

Your design software produces the layout, system size and production estimate. Revcore does not design solar systems, and neither should a chatbot. What AI can do is help your consultant prepare the conversation: turn the design output into plain-language talking points, write a one-line reason for each option, and draft answers to the objections this homeowner raised on the qualification call.

Options matter in solar as much as in any trade. One take-it-or-leave-it system turns the visit into a yes or no decision. Three honest options turn it into a which one decision. Our guide to good, better, best pricing covers how to build them without padding.

Once the options exist, the presentation itself should not live in a slide file someone rebuilds every week. In Revcore, interactive in-home sales presentations run on an iPad with your company story, the options side by side, and the homeowner's choice flowing into the estimate. The homeowner signs with an e-signature and pays the deposit at the table, by card or ACH, through Revcore payments. Presentations and Good/Better/Best start on the Pro plan.

How can AI keep homeowners updated between signing and permission to operate?

The quiet weeks after signing are where solar companies lose goodwill. The homeowner paid a deposit, nothing is on the roof, and the only way to learn what is happening is to call, which pulls a coordinator off the next job.

AI helps by drafting one clear message per stage, once, in your voice. The coordinator approves the set, and from then on each update goes out when the job actually moves. In Revcore, the homeowner portal gives the homeowner one place to see a phase-by-phase project timeline, photos, documents like the permit, and what they owe, and two-way texting, included on every plan, lets the coordinator send the approved update and read the reply in the same thread.

Keep AI away from dates it cannot know. A message can say the building department is reviewing the permit, never promise a turnaround you do not control. Install dates should come from the scheduling board the crew works from.

Can AI build permitting checklists for solar installs?

If you install across several cities or counties, every jurisdiction has its own forms, plan set requirements, fees and inspection steps. Coordinators often keep this in their heads or in a folder of old emails. AI is good at turning a building department's published requirements into a clean checklist. It is not good at knowing whether those requirements changed last month.

1. 1

   Start from the official source

   The permit coordinator downloads the building department's current solar permit page or checklist. AI works from that document, never from memory.

2. 2

   Ask for a checklist, not advice

   Prompt: turn this document into a numbered checklist of what we must submit, in order, with a column for who on our team owns each item. Do not add requirements that are not in the document.

3. 3

   Check line by line

   The coordinator compares every item to the source and fixes anything missing or invented. Code and structural questions go to the responsible engineer or designer. Date the list and recheck it after any rejection.

4. 4

   Save it where the team works

   Store the checked list in your shared templates for that jurisdiction and attach it to every job there. In Revcore, the job record keeps notes, photos and documents together, and permit documents can be shared in the homeowner portal.

How should solar installers use AI for reviews?

AI can draft a warm review request and a short, specific reply to each review. Ask when the system turns on and the homeowner sees it working, not the day the crew leaves a roof that is not yet producing. An automation tied to the final stage can queue the request for a person to approve.

Two limits. First, the FTC's rule on consumer reviews, effective October 21, 2024, bans fake reviews and bans offering incentives conditioned on a review expressing a particular sentiment, as its questions and answers page explains. Never ask AI to write reviews, and never tie a gift card to a five-star rating. Second, replies to negative reviews about production or billing should be written by a person who knows the job. An AI reply that guesses at the cause makes it worse.

Where does Revcore fit for a solar installer?

Revcore is not solar design software, and we would not tell you to replace your design tool. Revcore is the system that runs everything around the design: lead capture, the CRM and pipeline, estimates with Good/Better/Best options from your catalog, the iPad presentation, e-signatures and deposits at signing, scheduling and dispatch from one board, GPS clock-in and clock-out, job tracking with photos, the homeowner portal, invoicing, card and ACH payments with fast payouts, and reporting and marketing attribution. Flat plans include seats, and setup and migration are included on paid plans. When the lead, proposal, permit status and install date live in one record, AI drafts have the right facts to work from.

• Revcore AI Assistant. Included on every plan, now in beta. It is built to help with office work like spotting who owes money, which leads went quiet and what needs attention, and drafting follow-ups and reminders for a person to approve. Nothing goes out without approval. See Revcore AI.
• Connect your own assistant on Scale. On the Scale plan, you can connect Claude, ChatGPT or another MCP-capable assistant to your Revcore data with an API key. Ask about clients, leads, open estimates and invoices, active jobs, today's appointments, revenue and recent payments, and create clients and leads. It cannot create estimates. The setup is in our MCP guide.

If you are comparing systems, our roundup of the best solar installer software covers the field fairly, and the solar page shows how Revcore is set up for installers.

How do you roll out AI in a solar company in 30 days?

1. 1

   Week 1: set the rules

   Write a one-page AI policy. Customer data goes only into business or team AI plans with training turned off. No AI-written numbers, designs, savings or incentive statements. A person approves everything a homeowner sees. Name one owner, usually the operations manager, for the approved library.

2. 2

   Week 2: qualification and education

   Draft the qualification questions and one homeowner explainer. Test the questions on ten live leads and fix what confuses people.

3. 3

   Week 3: stage updates

   Draft the nine stage messages from signing to permission to operate. Load them as templates and have coordinators send them from the job record for every active job.

4. 4

   Week 3: permitting checklists

   Build checklists for your three busiest jurisdictions from official sources, checked by your permit coordinator.

5. 5

   Week 4: proposals and reviews

   Have consultants use the proposal prep prompt before each appointment, and set up the review request at permission to operate. At the end of the month, keep what worked and drop what did not.

Frequently asked questions

Further reading: AI for contractors in 2026, how to connect Claude or ChatGPT to your contractor CRM, what an in-home sales presentation is, how to follow up on an unsold estimate, and the best solar installer software roundup.