A cold snap triples the phone and your answer rate decides what you keep
Demand in this trade arrives in bursts tied to weather, and a shop staffed for an average week loses most of what a cold week sends it.
Track answer rate and missed calls by hour and by day, not as a monthly average. The pattern will show a spike on the first genuinely cold mornings and again after the heating bills land, and it will show exactly which hours go unanswered.
An unanswered call is not a lead you can nurture. Most homeowners simply dial the next result. Before testing anything on the website, fix the hours where nobody picks up.
Have a plan for the surge weeks: a second person on the phone, a spillover answering service briefed with your actual qualifying questions, and a rule that a voicemail gets a callback within a set window rather than whenever someone gets to it.
Ice events distort the mix. A share of that traffic wants emergency repair you do not sell. Give the front desk a one-line script that redirects those callers quickly so they do not consume the hour when whole-house buyers are also calling.
An estimator sent from Hendersonville to Franklin has already lost the morning
Measure appointments are your scarcest resource in a metro spanning five counties, and routing them badly costs more than any advertising inefficiency.
Group appointments geographically by day. Davidson on certain days, Williamson on others, the outer counties batched. Sumner and Wilson bookings should not be scattered across a week that also includes Rutherford.
Qualify enough at intake to know whether the drive is worth it. Address, opening count, what is failing, and whether the person on the phone is the decision maker. A two-opening inquiry in Murfreesboro deserves a different path from a whole-house inquiry in Belle Meade.
Build a cheaper path for the small jobs rather than declining them. A phone quote with photos, a minimum job size stated plainly, or a scheduled visit on a day you are already in that county. Turning them away entirely costs you referrals.
Watch the gap between inquiry and measure. Every extra day is a day the homeowner spends with a competitor who booked faster. Speed to the appointment is one of the few things in this business that reliably moves the close rate.
Qualify the openings before the appointment, not on the customer's porch
An estimator who arrives without knowing the window count, the frame condition or the house age is doing discovery on the clock.
Standardize the intake questions and hold everyone to them. Number of openings as a range. Approximate age of the house or the current windows. What is failing: drafts, fog between panes, operation, appearance, noise. Whether a date is driving the project.
Ask about the trim and the interior finish on older houses. A homeowner in Sylvan Park or 12 South with original wood sashes carries a different anxiety than a Mount Juliet owner with failed builder units, and the estimator should know which conversation to prepare for.
Collect a photo whenever possible. A single picture of the sill and the sash tells an experienced estimator more than five minutes of description, and it prevents the appointment where the answer turns out to be a repair.
Log the answers in the CRM in structured fields, not in a notes box. Free text is invisible to reporting, and reporting is how you find out which qualification question actually predicts a signed job.
The second and third bids arrive later, so follow-up has to outlast them
Assume the homeowner is collecting more than one quote, and that yours is being read next to somebody else's a week after you left.
Send the quote the same day when you can, itemized by opening rather than as a single number. A line-by-line document survives comparison. A lump sum invites the buyer to shop the number alone.
Build a cadence rather than a single follow-up. A call, then an email answering the objection the estimator heard, then a message addressing the difference between an insert and a full frame quote, because that difference is usually why a rival number looks lower.
Give the cadence a reason to continue past two weeks. Order lead times, a seasonal install schedule filling up, or simply an offer to walk through the competing quote. Do not manufacture urgency you cannot back up.
Newcomers need more of this, not less. A household that moved to Middle Tennessee recently has no basis for judging any of the bids, and the company that helps them compare fairly is frequently the one that signs.
A dashboard that stops at leads hides the two places you actually lose money
Counting inquiries tells you nothing, because the leaks live between inquiry and measure and between quote and deposit.
Instrument five stages: inquiry received, inquiry qualified, measure booked, measure held, contract signed. Report the drop between each. The two biggest drops in most window shops are qualified to booked, and quoted to signed.
Split every stage by county and by job type. A blended close rate mixing a whole-house Brentwood job with a single door in Donelson describes nothing that exists.
Count openings, not just jobs. One signed contract for eighteen windows and one signed contract for a patio door look identical in a job count and are nothing alike in revenue or crew time.
Reconcile against the calendar and the deposit, not the CRM stage. Salespeople advance stages optimistically. Money and scheduled install dates do not.
Sequence your changes by season, because the sample never arrives all at once
One metro at this volume will not produce a meaningful split test on a button, so spend the year testing things large enough to see.
Rank changes by expected effect: answer rate, speed to appointment, quote format, follow-up cadence, form length, then anything visual. Work down the list one item at a time and hold each in place long enough to read it.
Compare against the same period, not the previous month. A December to January comparison in a heating-driven market tells you about the weather rather than about your change. Year over year is imperfect and far more honest.
Change one thing at a time. If the form is shortened and the callback promise changes in the same week, you have learned nothing except that something moved.
Write down the baseline before you touch anything. Answer rate, inquiry to measure rate, measure to contract rate, average openings per contract. Without those four numbers, every later argument about what worked becomes a matter of opinion.
Questions we actually get
- What conversion rate should we expect from our website?
- Any agency quoting you a number before seeing your data is guessing. Rates vary enormously by traffic source, county, job type and how fast you answer. Measure your own baseline across a full season, then judge every change against that. Your own trend line is the only benchmark that means anything.
- Should we answer calls after hours in the winter?
- Look at your missed call log by hour before deciding. In a heating-driven market a real share of inquiries come in the evening from someone sitting in a cold room. If those hours show consistent volume, a briefed answering service or an on-call rotation usually pays for itself faster than additional advertising does.
- Is an answering service a good idea for a window dealer?
- It depends entirely on the brief. A service that takes a name and number and promises a callback beats voicemail. A service that gets the address, the opening count and what is failing is far better. Give them your actual qualifying questions, listen to recordings monthly, and drop them if the transcripts show inquiries handled worse than voicemail would.
- How do we handle small jobs without wasting an estimator's day?
- Give them a separate path rather than a rejection. A photo-based phone quote, a stated minimum, or an appointment slotted on a day you are already working that county. Small jobs in Hendersonville or Murfreesboro become expensive only when they are routed the same way as a whole-house project in Green Hills.
- What should we test first if we can only change one thing?
- Speed. Time from inquiry to first human contact, then time from inquiry to booked measure. In this trade both correlate more strongly with signed work than anything on the page does, and both are free to change. Save form fields, headlines and layout for after those two numbers are where you want them.
