AI for Solar Companies: Qualifying Solar PV Leads Automatically

Many solar businesses have less of a demand problem than a sorting problem. Enquiries come in through the website, by phone and from comparison portals, but only some of them lead to a site visit. This article shows how to pre-qualify solar PV leads with AI, which questions are enough and where expert advice remains essential.

Why solar businesses lose so much time on unsuitable enquiries

A typical Monday morning at a regional PV installer: twelve enquiries came in through the contact form over the weekend, plus messages on the answering machine and several leads from a comparison portal. Before anyone knows which are worth pursuing, the office has to call them all back. Some can't be reached, some are renting, others have a north-west roof under old trees, and some only wanted to know a price.

Speed matters especially with bought portal leads, because the same enquiry often goes to several providers. If you only get in touch on Wednesday, you've often already lost. At the same time, the advisers who should be working out quotes and inspecting roofs are busy returning calls. Here a digital assistant can handle the first contact and separate the wheat from the chaff before a person invests time.

Qualifying solar PV leads: the questions that really count

A first assessment doesn't need twenty questions. Most businesses manage with six to eight points that an AI assistant can clear up in a few minutes in a chat, on WhatsApp or on the phone:

• Ownership: does the enquirer own the building, or would a landlord or the owners' association have to agree?
• Building and roof: detached house, apartment building, commercial hall; roof shape, rough orientation, age of the roof covering.
• Location: is the address within your installation area?
• Power consumption: approximate annual consumption from the electricity bill, plans for a heat pump or an electric car.
• Scope: modules only, with battery storage, with a wallbox?
• Timeframe: a concrete plan for the coming months, or still gathering information?
• Photos: optionally, prospects can send pictures of the roof and the meter cabinet for the adviser to look at before the appointment.

What matters is what happens with the answers. If everything fits, the assistant offers a consultation slot straight from the responsible adviser's calendar. If the enquirer is a tenant, they can be pointed politely to alternatives such as a balcony solar kit, instead of simply getting no reply. And anyone just gathering information receives material and a follow-up later. You can read more about the basics in the article on automated lead qualification.

Worked example for a regional installer

The following figures are assumptions for illustration, not measurements. Plug in your own numbers.

Suppose a business with two advisers receives 120 enquiries a month. So far the office calls every enquiry back, taking 15 minutes on average including several attempts. We assume that 40 enquiries turn out to be unsuitable in the first conversation and that another 30 have no concrete purchase date. On top of that, around six site visits a month used to take place where it only became clear on the roof that a system wasn't a fit. Each of these visits costs around two hours including travel.

Item	Assumption	Time per month
Callbacks for unsuitable enquiries	40 enquiries at 15 minutes	10 hours
Callbacks for information-only enquiries	30 enquiries at 15 minutes	7.5 hours
Avoidable site visits	6 visits at 2 hours, assuming half could have been spotted through questions and photos	6 hours
Total		23.5 hours

Whether this time really becomes free depends on how well the pre-qualification works and how many enquirers complete the conversation with the assistant. So also run a cautious version with half the values. If it still pays off then, the decision is solid.

How to introduce pre-qualification

1. Define exclusion criteria. Agree as a team on the few points where an enquiry definitely won't fit, such as installation area or lack of ownership.
2. Phrase questions in everyday language. Not everyone understands "What is the orientation of your roof?". "Roughly which direction does the largest roof surface face?" works better.
3. Bring channels together. Website form, phone, WhatsApp and portal leads should feed into the same process, so no source gets left lying.
4. Read along for four weeks. Check which rejected enquiries the adviser would actually have accepted, and adjust the rules. Ideally, sit down with the adviser for half an hour once a week and go through a selection of conversation logs together.

Following up after the quote

Many PV quotes don't fail on price but because customers compare several quotes and keep putting off the decision. A friendly follow-up after a few days, asking about open questions and offering a phone call with the adviser help more here than discounts. These steps can be automated, as long as the customer has agreed to be contacted. The article on automatic follow-up on quotes shows concrete processes. If the customer asks about financing, subsidies or tax implications, the assistant should hand over to the adviser and not make any statements of its own.

Limits of automated pre-qualification

An assistant can't judge roof statics, measure shading or inspect a meter cabinet. It collects the details the customer gives, and these are sometimes wrong. The binding assessment stays with the expert adviser. Nor should the AI promise yields, payback periods or savings. Such statements depend on many factors and can cause trouble later if they create false expectations.

A filter that is too strict is a risk too. If you reject every enquiry with an east-west roof, for example, you lose customers for whom a system can make perfectly good sense. Better to start with a few hard exclusion criteria such as installation area and ownership, and let the adviser check borderline cases. With portal leads, also check whether the prospect's consent covers calls and messages from your business.

Frequently asked questions

Do prospects mind speaking to an AI first?

Most expect a quick response and take an honest notice about the AI assistant in their stride. What matters is that the notice comes at the start, as the EU AI Act also requires, and that the customer can ask for a callback from a person at any time.

How do I rate leads that don't clearly fit?

With a simple points system your sales team understands. Ownership and installation area are mandatory; timeframe and scope determine priority. The article on AI lead scoring explains how such ratings are built.

Can the assistant also schedule appointments around travel times?

It can put appointments in the responsible adviser's calendar while following fixed rules, such as buffers between site visits or certain days for certain regions. It doesn't replace proper route planning.

Conclusion

If you pre-qualify solar PV leads quickly and against clear criteria, your advisers get time for the conversations that can lead to orders. AI takes over the first contact, the standard questions and the scheduling. The technical assessment and any statements about economics and subsidies stay with people.

Neurobots uses digital employees for this that take enquiries by phone, website chat and WhatsApp, book appointments and pass qualified contacts to CRM systems such as HubSpot or Pipedrive. The page Digital assistant for trades businesses gives an overview for specialist firms.