The morning after a hard freeze, more work is lost at the phone than in the field
When demand spikes past capacity, the only variable you control is what happens to the calls you cannot serve today.
Freezing weather produces a volume of no-heat calls no shop can absorb. Most of those calls end with an apology and a hangup, which means paying to acquire nothing and giving the customer to the next number on the list.
Build a triage script before winter. Offer the next real appointment rather than a vague callback. Take the details anyway. Give a short safety note about what to do until somebody arrives, and offer to move them up if a slot opens.
A caller who is booked for Thursday is still yours. A caller told you are slammed is not. The difference costs a minute of the phone's time and no marketing spend at all.
Track how many calls in a surge week ended without a booking. Nobody counts it, and in most shops it is the single largest recoverable number in the business.
Nobody in the office has listened to what the after-hours service promises at one in the morning
Your answering service is quoting your capacity, your pricing posture and your response time to every emergency caller, unsupervised.
Pull twenty recordings from the last cold snap and listen. The common failures are consistent: promising a callback nobody made, quoting an arrival window the board cannot support, or taking a message with no address and no phone number spelled back.
Write the script yourself. What they may promise, what they must ask, what they must never quote. Include the fuel type question, the property address and whether the caller owns, rents or manages the property.
Set a rule for the morning: after-hours messages get returned before anything else, in the order received. A call returned at ten has usually already booked somebody else.
If the service cannot follow a script, the money spent on it is buying a worse outcome than a well written voicemail with a clear promise attached.
Five steps sit between a ring and a signed job, and most shops count one
Lead totals tell you nothing about where the work is falling out, because the loss almost never happens where people assume.
The path is answered, booked, arrived, quoted, sold, and for replacements, financed. Each of those is a number. Most companies track the first and the last and have no idea what happens between them.
Once the numbers exist the diagnosis is usually obvious and rarely what the owner expected. Booked to arrived leaks in bad weather. Quoted to sold leaks when quotes are delivered verbally. Answered to booked leaks at specific hours nobody has looked at.
Split the same numbers by county. Davidson County and Williamson County frequently behave differently in the sold column, and the reasons are worth knowing before you change anything about the marketing.
Reconcile against the dispatch board weekly, not monthly. A month is long enough for everyone involved to have a plausible story about what happened.
A replacement quote in Brentwood dies in the week nobody follows up
System replacement is a considered purchase, and considered purchases are won in the gap between the visit and the second conversation.
The technician leaves, the homeowner talks it over, and the second and third contractors arrive. If your next contact is a text three weeks later, you are quoting against a decision that has already been made.
Set a sequence and keep it. A written proposal the same day. A call inside two days from somebody who can answer questions the technician could not. A final check before the quote expires. Nothing elaborate, just reliable.
Give the proposal something to stand on: what each option includes, what it does not, what the difference actually buys, and how financing works if you offer it. Newcomers with no contractor to compare against read that document more carefully than anyone else.
Count the proposals that expired without a second conversation. In most shops that number, multiplied by an average job, is larger than the entire marketing budget.
The technician in the crawlspace is the one who can sell the maintenance agreement
Agreements are what turn a one time panic call into a customer, and they are sold at the job, not by email afterward.
Somebody who has just watched a technician find and fix the problem is at the highest point of trust they will ever have with your company. Two weeks later they are a name in a database.
Give the technician a simple offer, a simple explanation and a way to sign it on a phone before leaving. Every extra step between the pitch and the signature costs conversions.
Track agreement rate by technician. The spread is usually wide, and the top performer's approach is a training program you already own and are not using.
Agreements also change what you can afford to pay for a first call. A customer on a plan does not open a search engine next winter, which is the entire point.
Weather wrecks before-and-after comparisons, so change big things and hold them a season
One metro does not produce enough traffic for small split tests, and a cold snap will manufacture whatever result you were hoping for.
Compare two months in a market like this one and the weather explains more of the difference than anything you changed. Small button and headline tests are unreadable at this volume, and reading them anyway is worse than not testing.
Test things large enough to survive the noise. The after-hours answering arrangement. Whether proposals are delivered in writing the same day. Whether the form asks for fuel type. A new manager focused landing page. Changes with a mechanism you can explain.
Hold each change long enough to see a real sample, and where possible compare like periods, this January against last January, rather than January against October.
Some improvements do not need a test at all. Answering the phone during the two hours you currently miss most calls does not require statistical proof, only a schedule.
Questions we actually get
- Where should we look first if we only have time for one thing?
- The phone during your busiest hours and the after-hours answering arrangement. Pull the call records for the last cold snap and count the calls that ended without a booking. In most HVAC companies that single number is larger than anything conversion work on the website could recover, and fixing it costs staffing decisions rather than budget.
- How do we know whether a change worked?
- Compare like periods and prefer large changes. This metro does not generate enough volume for reliable small tests, and weather swamps short comparisons. Set a baseline for booked rate, sold rate and agreement rate before the change, hold the change through a comparable stretch of season, and look at the same numbers by county afterward. Anything that only shows up in a two week window should be treated as noise.
- Should we quote prices over the phone?
- Give a range for diagnostic visits and be clear about what could change it. Do not quote a repair or a replacement without seeing the equipment. Callers who are new to the area often ask because they have no reference point, and a straight explanation of how pricing works satisfies them better than a number you would have to walk back.
- Our reviews mention response time more than price. What does that tell us?
- That your conversion problem is operational rather than commercial. Response time complaints usually trace to one of three places: after-hours messages returned late, arrival windows promised beyond capacity, or no communication when a technician runs behind. All three are process fixes, and all three affect booked to arrived rates before they ever affect reviews.
- Is it worth optimizing for property managers separately?
- Generally yes if you already do that work, because the buying process is different. A manager handling short-term rental properties around The Gulch or East Nashville is buying turnaround and predictable invoicing rather than a relationship, and often controls several properties. That means a separate intake path capturing property, access method and billing contact, and standing authorization limits agreed in advance so nobody waits a day for approval on a small repair.
