The honest answer about tonight books more work than the optimistic one
A caller with no heat is deciding between waiting for you and calling somebody else, and vagueness sends them to the somebody else.
When the board is full, the instinct is to soften it. We will try to get someone out today. The caller hears a maybe, hangs up, and dials the next result while your dispatcher is still typing.
A firm answer holds better. Tell them the truth: the earliest is tomorrow between eight and eleven, here is what to do tonight, and we will call if something opens. People accept a real answer far more often than an evasive one.
Give the phone team a script for the day the board is full. Include what to say about safety in general terms, and when to advise calling the gas utility or emergency services instead of waiting for a truck.
Then confirm in writing. A text with the window, the technician's name and your number reduces the cancellation that happens two hours later when a competitor calls back.
Your booking rate is three different numbers, one per channel
Phone, web form and text behave completely differently, and a single blended rate hides which one is broken.
Calls answered live book at one rate. Forms submitted at nine at night book at another. Texts are different again. Averaging them produces a number that describes nothing you can act on.
Split the reporting. Count inquiries, contacted, booked and completed for each channel separately, weekly. The channel with the worst drop between inquiry and contact is where the money is.
In most shops the form is the weak one, because nobody owns it after hours and it is treated as less urgent than a ringing phone. It is not less urgent. It is the same person who could not get through.
Set a target response time per channel and post it where the office can see it. Attention alone moves the number before you change any system.
The half hour after a technician condemns a furnace decides the month
The handoff from diagnosis to replacement quote is the highest value thirty minutes in the business, and it is usually improvised.
A cracked heat exchanger in January is a decision made under pressure in a cold house. Whoever handles the next thirty minutes competently tends to get the job.
Build the handoff. The technician documents the condition with photos, explains options in plain language, and either produces a same visit quote or gets a comfort adviser on the phone before leaving the driveway.
Leave something behind that survives the conversation. A written summary with the equipment details, the options and what happens next. It is what the customer shows a spouse that evening, and it is what a second contractor has to argue against.
Track how many condemned systems produce a quote, and how many quotes produce a sale, as two separate numbers. Most shops track neither and assume the loss happened somewhere else.
Sold work stalls at scheduling, not at the signature
The days between a signed proposal and an install date are where customers quietly change their minds.
A signature is not revenue. Between it and the install sit equipment availability, crew capacity and permitting, and every day in that gap is a day for second thoughts.
Permitting and inspection practice varies between municipalities in St. Louis County, and the city operates under its own rules as an independent city. Details are worth confirming with each local building department rather than assumed, because a scheduling promise made on the wrong assumption becomes an angry phone call.
Give the customer a named contact and a date for the next update, even when the news is that nothing has moved. Silence is what gets read as trouble.
Measure days from signature to install as its own metric. It is unglamorous and it correlates with cancellation more closely than most owners expect.
A missed call with no text back is the cheapest repair in the funnel
Every abandoned ring is demand you already paid for, and the fix is a message rather than a hire.
Missed calls cluster. Usually early morning after a cold night, mid afternoon during shift change, and any hour during a freeze event. Pull the phone records and the pattern will be obvious within an hour.
An automatic text back to any missed call, sent within a minute, recovers a share of them. Keep it short, name the company, and invite a reply describing the problem. Somebody has to actually watch the replies for it to work.
Then fix the clustering. Overflow answering during known peak hours costs less than the jobs lost, and it is easier to justify once you can see the pattern in your own data.
After hours matters most in heating season. Listen to what your answering service actually says at one in the morning before you assume it is doing what you asked.
Speed to answer moves faster than any split test you could run here
One metro does not produce enough traffic for small tests, so change the things big enough to see without one.
Split testing a button color needs volume that a single market rarely supplies, and heating demand swings with the weather badly enough to poison any before and after comparison.
Work on the changes large enough to show up regardless. Answering the phone within three rings. Returning every form within the hour during business time. Quoting on the visit rather than the following week. Effects of that size do not need a statistical test to be visible.
When you do compare, compare like periods. The same weeks against the same weeks a year earlier, with a note about what the weather did. Comparing January to October measures the season, not your work.
And listen to recordings. Reading a transcript of your own intake calls for one afternoon usually produces a longer improvement list than a quarter of testing.
Questions we actually get
- We answer nearly every call. Why is our booking rate still low?
- Answering is not booking. The common causes are a vague answer about timing when the board is full, no written confirmation after the call, and no follow up when a caller says they need to check with a spouse. Pull twenty recordings from your busiest week and the pattern usually appears in the first few.
- Should we use an after hours answering service?
- Often yes during heating season, with one condition: somebody in your office should call it at one in the morning and listen to what it actually says. Many services promise things you would never promise, or take a message without conveying any urgency. Audit it before the first freeze, not after.
- Does charging a diagnostic fee hurt conversion?
- It filters, which is usually helpful. The bigger factor is whether the fee is stated clearly before the caller asks and whether it applies to the repair. Hidden fees produce cancellations at the door, which cost far more than the calls a published fee turns away.
- How do we know which marketing produced a booked job?
- Somebody has to reconcile the dispatch board against lead source on a regular schedule, weekly at first. Call tracking numbers and a source field in your CRM get you most of the way, but only if the office is trained to fill it in and the reconciliation actually happens.
- Can we A/B test our website?
- Rarely with enough volume to trust the result, and seasonality makes it worse. Focus instead on changes large enough to see without a test: response times, quoting on the visit, follow up on open estimates. Compare the same weeks year over year, and note what the weather did before drawing conclusions.
