Qualifying Window Replacement Leads Before You Drive Out

The web form says: "Need price for new windows. Please call." There's a name and a phone number. After two rounds of phone tag, a sales rep books an in-home consultation and drives forty minutes to get there. Inside, it turns out the homeowner wants one bathroom window replaced, is thinking about "maybe next year", and the spouse who makes the decision isn't home.

That visit cost the company two hours of a salesperson's day and the fuel to get there. Multiply by a few wasted visits a week and it's a real line in the budget, not to mention the good leads that waited longer because the calendar was full of weak ones.

Window replacement is a business where the first visit is expensive and the information needed to decide whether to make it is easy to get. That's a good place for AI: not to replace the consultation, but to make sure the consultation is worth driving to.

The market context in 2026

Two facts shape how homeowners approach window projects right now.

With the federal credit gone, more homeowners want a sense of cost before they let anyone into the house. At the same time, the energy argument is solid, especially for older single-pane windows. A company that can give an honest range early, and explain what the homeowner is likely to gain, earns the consultation instead of begging for it.

What the form says versus what you need

Here's the gap, laid out.

None of the right-hand column is secret or hard to answer. Most homeowners can provide all of it in five minutes from their phone. They just aren't asked, or they're asked on the phone by someone reading a script, which is slower for everyone.

The flow I'd build

The step that makes the most difference is the second one. A reply within minutes, asking good questions, catches the homeowner while they're still thinking about windows. The same questions asked two days later in a phone call often get "I'll get back to you".

The first reply

The tone of that first message decides whether homeowners answer. It should sound like a helpful person at a local company, not a form. Something like this, adapted to what the homeowner wrote:

Hi Karen, thanks for getting in touch about new windows. So we can give you a realistic price range before anyone visits, could you tell us:

1. Roughly how many windows, and in which rooms?
2. What's bothering you about the current ones: drafts, fogging, noise, or just their age?
3. Do you own the house, and is there an HOA?
4. When would you ideally like the work done?

If it's easier, you can [upload a few photos here] instead, and we'll count from those. Prefer a call? Just reply with a good time.

Four questions, all answerable from the sofa. The photo option takes the place of the question most people find hardest, which is describing their windows. And the last line matters: a homeowner who wants to talk to a person should never feel trapped in an automated exchange.

Getting good photos

Homeowners are happy to send photos if you tell them exactly what to take. This short guide is what I'd put behind the upload link.

The tape measure photo matters. It won't replace a proper measurement, but it lets the range be realistic instead of a guess.

What AI can and can't see in a photo

It's worth being honest about this, because it's easy to oversell.

The lead paint row is not a detail in the US. In homes built before 1978, window replacement disturbs painted surfaces, and the EPA's Renovation, Repair and Painting rule requires certified renovators and lead-safe work practices. Asking "when was the house built?" in the first reply lets you plan for that before the consultation, not during it.

Should you give a price range before the visit?

I think yes, and this is where many window companies disagree with me. The fear is that a range scares people off. In my view, it scares off the wrong people: homeowners who wanted one cheap window, and homeowners who would never have signed. The right people appreciate knowing the ballpark before they give up an evening for a consultation.

The range must come from your own price book, not from a model's imagination. A rule like "vinyl double-hung, standard size band, installed: this range" multiplied by the counted windows, with a margin for unknowns, is simple and honest. The message says clearly that the final price follows a measurement.

Sorting leads by readiness

That's roughly eleven weeks of a salesperson's time, spent in living rooms that were never going to turn into orders. Even halving it frees capacity for the leads that are ready. The sorting itself is simple:

• Survey-ready: owner, clear scope, photos, timeline within a few months. Book the consultation.
• Needs one answer: almost there. The assistant asks the missing question.
• Later: real interest, but "next spring". A check-in at the time they named, not a weekly drip.
• Not a fit: a single glass unit to replace, or a job you don't do. A quick, friendly referral to a glazier or another trade.

What changes at the consultation

When a lead arrives prepared, the consultation itself gets better, not just rarer. The rep knows there are eleven windows, that three of them are fogged, that the house was built in 1962, and that the homeowners care most about noise from the road. They bring the right samples, including an acoustic glazing option, and plan for lead-safe work practices. Both decision makers are there because the booking confirmation asked for it.

The homeowner notices too. The conversation starts from their situation rather than from a generic pitch, and the price they hear at the end sits within the range they already saw. That consistency is what turns a consultation into a signed order: no surprises, no pressure, and a company that clearly listened before it showed up.

Tools and data

Most window companies run a CRM such as HubSpot, Salesforce, MarketSharp, Leap, or JobNimbus. The intake assistant creates or updates the lead there, attaches photos and the summary, and sets the stage. Your price book can live in a spreadsheet to begin with; it just has to be the one source both the assistant and your reps use.

Homeowners share photos of the inside of their homes, which deserves care: store them with the lead, limit access to the sales team, and delete them for leads that don't go anywhere after a sensible period. In the EU, the GDPR applies, and people should be told when they're chatting with an AI under Article 50 of the AI Act.

Questions window companies ask

Won't asking questions up front scare leads away?

Some leads will drop off. Those are mostly the ones that would have been wasted consultations. Keep the questions few, conversational, and optional, and always offer a phone call for people who prefer it.

Can AI measure windows from photos?

Not accurately enough to order. With a tape measure in the frame it can estimate a size band, which is good enough for a range. Ordering always needs a proper measurement by your installer.

Should we put prices on the website instead?

A price guide on the website helps, but it can't account for the number and types of windows in a specific house. The assistant's range is more useful because it's based on what this homeowner actually sent you.

What if a homeowner won't send photos?

Then the lead stays in "needs one answer" and a person calls. Some people just prefer talking. The point isn't to force everyone through a form, it's to stop driving out blind.

The rule of thumb

Drive out for a measurement, not for an introduction. Everything a homeowner can tell you from their phone should be known before anyone gets in the car.

If your sales calendar is full of consultations that go nowhere, tell me how leads reach you today and which CRM you use. I'll suggest a small intake assistant you could test within a month. The same logic applies to solar installers deciding which site visits to make, and to security installers preparing site surveys.