Two questions on the first call separate a whole house from one stuck door
The first ninety seconds should establish opening count and the age of the house, because everything downstream depends on both.
Opening count decides whether an in-home measure pays. Fourteen openings justifies the drive from your yard to almost anywhere in the Metroplex. One patio door in another county usually does not, and pretending otherwise is how estimators lose days.
House age tells you the product path before anyone quotes. A house from the mid-century core means non-standard openings, matching questions and slower work. A house from the growth years in Frisco or McKinney usually means failed seals in builder units and a faster path to a whole-house number.
Write both into the intake script as the first two questions, and make sure whoever answers the phone knows why. A script that reads like an interrogation loses people; two questions asked conversationally do not.
Add a third question for anything that sounds structural. If the complaint is a door that suddenly stopped latching or a gap that opened at a corner, note it before booking, because that appointment needs a different conversation and sometimes a referral to a licensed engineer rather than a sale.
Denton County measures should stack on one day instead of scattering across three
Appointment scheduling is a conversion lever disguised as an operations problem.
An estimator who drives from Carrollton to Denton, back to Garland, then out to Grapevine has spent a day producing three quotes. The same day scheduled by corridor produces five or six, and every additional slot is a job you were previously turning away.
Build the booking system around geographic days. Northern corridor on certain days, Dallas core on others, Tarrant County on others. Publish the available slots that match, rather than offering an open calendar and sorting it out later.
The catch is speed. A homeowner who has to wait nine days for the corridor day will call someone else. Keep a small number of flexible slots for large-opening-count inquiries so the valuable jobs are never the ones waiting.
Watch what corridor scheduling does to your close rate by area. If Tarrant County closes materially worse because appointments are always a week out, that is a business decision about whether to serve it properly or not at all.
Storm damage inquiries need their own track and a script that stays out of the claim
A hail-driven inquiry follows a longer, more paperwork-heavy path than a retail one, and running both through the same pipeline damages both.
Severe hail produces a replacement market that runs through insurance rather than a discretionary budget. The buyer behaves differently, the timeline is longer, and there is a third party in the middle of the decision.
Separate the track at intake. Different questions, different follow-up cadence, different expectations about when a signature happens. Blending them makes your retail close rate look worse than it is and your storm follow-up look faster than it is.
Keep the script strictly on your side of the line. Describe what you install and what you document. Do not advise on coverage, do not offer to handle a claim, and route the homeowner to their own insurer for anything about the claim itself. Texas regulates what a contractor may say here, and it is worth having your own counsel review the script and any related paperwork.
There is a conversion benefit to that restraint. Homeowners in the Metroplex have met the crews who show up after a storm promising outcomes. Being the company that will not overreach, with a permanent address and a warranty, is a position competitors cannot copy quickly.
Your close rate is three different numbers hiding under one average
A single close rate across a mixed pipeline tells you nothing you can act on.
Segment it at minimum three ways: by opening count band, by submarket, and by inquiry source. A dealer closing well on whole-house work in Plano and badly on two-opening inquiries in Mesquite has an average that describes neither situation.
Then segment by estimator. Variation between people is usually larger than variation between marketing channels, and it is the cheapest thing to fix because it responds to training and a shared quote format.
Record calls where the law and your policy allow it, and listen to a sample every week. Call recordings surface the same three objections repeatedly, and once you can name them you can answer them in the quote document instead of hoping someone remembers on the phone.
Track through to signed work, not to quotes issued. A team can produce quotes all week without producing revenue, and a report that stops at the quote will happily celebrate it.
Canceled measures, not missed calls, are where most Metroplex dealers bleed
The appointment that was booked and never happened is the most expensive event in this business, and almost nobody counts it.
Count no-shows and cancellations as a distinct metric with its own number. If a meaningful share of booked measures never occur, every improvement upstream is being poured into a leaking tank.
Confirmations are the cheap fix. A same-day text with the estimator's name, a photograph of the person arriving, and a simple reply option to reschedule rather than silently cancel. People do not cancel because they lost interest; often they cancel because something moved and rescheduling felt like a phone call they did not want to make.
Shorten the gap. The longer between booking and appointment, the more the appointment decays, which is another argument for keeping flexible slots for high-value inquiries.
Give the homeowner something to do in between. A short note on what the measure covers, how long it takes, and whether everyone who makes the decision should be there converts better than silence, and it prevents the appointment where the spouse is missing and nothing gets decided.
A metro this size funds a few real tests a year, so make each one an offer change
There is not enough volume in one metro to detect small effects, which means testing small things is a way of learning nothing slowly.
Button colors and headline variants need traffic that a regional dealer does not have. Running them produces results that look conclusive and are not, and the account spends a quarter chasing noise.
Test things large enough to move the number on their own. The offer itself, the intake script, the appointment confirmation sequence, the structure of the quote document, whether a photo upload appears on the form. Changes of that size show up in booked measures without statistical gymnastics.
Run them sequentially and hold everything else steady, comparing full periods against each other rather than partial weeks. Note the season, because demand rises ahead of the hot months and after severe weather, and a change made in a busy week will look better than it is.
Measure in booked measures and signed openings, never in form submissions. A form change that raises submissions and lowers appointment quality is a loss dressed as a win, and the only way to see it is to follow the number all the way through.
Questions we actually get
- Where do most window and door inquiries actually leak?
- In most shops it is not the website. The largest losses are unqualified bookings that waste an estimator's day, appointments that are booked and then never happen, and follow-up that stops after one attempt on a purchase homeowners take weeks to decide. Instrument those three stages first and the biggest gap usually becomes obvious within a month.
- How fast do I need to answer an inquiry?
- Faster than the alternatives, and the alternatives are a search results page with several other dealers on it. Answering live during business hours matters most; after hours, an immediate acknowledgment that sets a specific callback time holds people better than silence. What matters more than any specific target is that you measure the delay at all, because most shops have never looked.
- What should I qualify before booking an in-home measure?
- Opening count, address or at least zip code, approximate house age, and whether the complaint sounds structural. Across a metro this size those four answers tell you the drive, the product path, the likely ticket, and whether this is a window job at all. Add whether everyone who makes the decision will be present.
- Can I run A/B tests with the traffic a single metro produces?
- Not for small effects. There is rarely enough volume to detect a modest change with confidence, so testing button copy or color mostly produces noise. Run sequential tests on large changes such as the offer, the intake script, the confirmation sequence or the quote format, compare full periods, and note the season, since demand shifts ahead of the hot months and after severe weather.
- How should I handle inquiries that turn out to be foundation problems?
- Catch them at intake by asking about doors that suddenly stopped latching or gaps that appeared at corners, then have a different conversation. Explain what movement does to an opening and what a replacement unit can and cannot correct, avoid diagnosing the property, and refer the structural question to a licensed engineer. Some of those homeowners come back once the slab issue is settled.
