Book the inspection on the first call or admit you cannot
Every hour between an inquiry and a scheduled appointment is an hour a competitor uses.
The person calling about a leak has a short list and is working down it. The person planning a replacement is more patient but still forming an impression in the first thirty seconds. In both cases the outcome you want from that call is not a name in a system. It is a date.
Give whoever answers the ability to place an appointment. If the calendar has to go through a manager, it will not happen the same day, and the inquiry ages. Aging inquiries close at a fraction of the rate.
If you genuinely cannot book on the call, be explicit about the next step and hold to it. A specific commitment beats a vague one every time, and a call-back window that is met builds more trust than a promise that is not.
Measure the gap. Time from inquiry received to appointment scheduled, tracked weekly. It is the single most useful operational number in this trade and almost nobody records it.
Handle after-hours honestly. Either somebody answers or the voicemail says exactly when a person will, and then a person does.
Route the estimator by zip code cluster, not by who called first
In a metro where drive time decides everything, first-come scheduling quietly cuts the number of roofs anyone sees in a day.
An estimator who takes Alpharetta at nine, Decatur at eleven thirty and Marietta at two spends most of the day on the interstate. The same person working one cluster sees twice as many properties and arrives less irritated.
Cluster geographically and by half-day. Inside the Perimeter in one block, north Fulton and Cobb in another, the Gwinnett side in another. Tell callers the day you run their area rather than offering an open calendar, and most will accept it.
Give an arrival window and text when the estimator leaves the previous job. Missed and rescheduled visits are one of the largest silent losses in roofing, and most of them come from a window nobody could keep.
Confirm the day before. Ask about gate access, dogs, whether a second decision maker will be present, and whether anything overhangs the roof. Every one of those has ended a wasted trip before it started.
Count completed visits rather than booked ones. The gap between the two is where the routing problem shows up.
Most lost roofs are not lost, they are waiting on a spouse and an adjuster
The proposal is not the end of the sale, and treating it that way is why healthy pipelines go quiet.
Replacement decisions here involve two people, sometimes a claim, sometimes a homeowners association color approval, and often a wait until after a large expense elsewhere. None of that is a no. All of it looks like a no if nobody follows up.
Build a sequence and run it whether or not anybody feels like it. A call the day after the proposal, a message a few days later answering the objection you actually heard, something useful at two weeks, and a genuine check-in before the next season. Written down, assigned, and visible.
Follow up with substance rather than a nudge. Photos from the inspection. A plain explanation of what happens if the repair waits another year. An answer to the specific question they hesitated on. Anyone can send a reminder.
Keep claim conversations careful. Document what you saw and what it will take to fix, and let the homeowner and their insurer decide what is covered. Do not tell somebody a claim will be approved.
Track the reasons. If half your stalled proposals are waiting on a second decision maker, the fix is at the inspection, not in the follow-up.
Count contracts per inspection, and split the number by county
A blended close rate across a metro this fragmented describes a market that does not exist.
Set up the funnel properly: inquiry, appointment scheduled, appointment completed, proposal delivered, contract signed, job installed. Each step has a drop, and the biggest one is almost never where the owner assumes.
Split those steps by area. Intown work and outer suburb work behave differently at nearly every stage, from how many appointments are kept to how long a decision takes. One number hides both.
Split by job type as well. Repair and replacement do not belong in the same close rate, and mixing them makes a good repair month look like a bad sales month.
Attribute back to source where you can, and look back rather than only at the current month. Roofing cycles are long enough that the click and the contract regularly land in different reporting periods.
Keep the reporting small enough that somebody actually reads it. Six numbers reviewed weekly beat a dashboard opened twice a year.
There is not enough traffic here to test a button, so test the process
One metro in one trade will not produce the sample size a small-effect split test needs, and pretending otherwise produces confident wrong answers.
A test that needs thousands of conversions to detect a two percent difference is not going to resolve on your traffic. Running it anyway means reading noise and shipping changes that do nothing.
Test things large enough to see. Whether a phone-answering script that opens with the address changes the booking rate. Whether a day-before confirmation cuts missed visits. Whether offering a repair path changes how many inquiries become anything at all. Those are big effects, and big effects show up in small samples.
Change one thing at a time and hold it long enough to cover a full cycle of your demand, including the slow weeks. Seasonality here runs twice a year, so a two-week read is not a read.
Use qualitative evidence more aggressively than a high-traffic business would. Ten recorded calls and five conversations with customers who did not buy will tell you more than an underpowered test, and cost less.
When the sample really is too small, say so out loud rather than dressing a guess up in a percentage.
The quiet weeks in January are the ones worth staffing properly
Two demand seasons mean predictable slow stretches, and what you do in them decides how the busy ones go.
Unlike markets with one long season, this metro has genuine winter demand and genuine winter quiet. Volume drops, and the instinct is to cut coverage. The inquiries that do arrive in those weeks are often better qualified, because nobody idly researches a roof in the cold.
Use the slow period for the work that never fits otherwise. Call the stalled proposals from the fall. Clean up the intake notes. Sit with the office and listen to calls together. Fix the confirmation email nobody has read since it was written.
Book ahead. An inspection in January and an installation in March is a good outcome, and it smooths the spring rush that otherwise causes the missed appointments and slow callbacks you spend the rest of the year fixing.
Set the staffing to answer, not to match volume. A missed call in a quiet week costs proportionally more, because there are fewer of them.
Questions we actually get
- What is the single biggest leak in a roofing funnel?
- In our experience with the trade generally it is the gap between an inquiry arriving and an appointment being on the calendar, followed closely by appointments that get booked and never happen. Both are operational rather than marketing problems, and both are measurable this week without changing anything on your website. Start by recording time to first contact and the ratio of completed visits to booked visits, then look at which is worse.
- How much traffic do we need before split testing is worth it?
- More than a single metro in a single trade usually produces, if you are testing small changes. The honest position is that most roofing companies cannot detect a small difference reliably, so the effort belongs on changes large enough to show up anyway: intake scripts, booking policy, confirmation habits, whether a repair path exists at all. When somebody shows you a test result claiming a small improvement on a few dozen conversions, be skeptical of it.
- Should we use a call answering service?
- It depends entirely on whether the service can book an appointment. A service that takes a message adds a step and slows the thing that matters most. One that has access to your calendar, can ask the address first and can schedule inside your routing clusters can help a great deal, especially after hours. Test it by listening to the recordings for a month rather than by reading their reporting.
- How long should we chase a proposal before giving up?
- Longer than most companies do. Roofing decisions here often wait on a second decision maker, an insurance process, an association color approval or a budget cycle, and none of those resolve in a week. A sequence that runs across several weeks and then checks back before the next season costs very little and recovers work you have already paid to generate. Stop when someone tells you to, and record the reason.
- What should we measure if we can only track a few things?
- Five numbers. Time from inquiry to appointment scheduled. Completed visits as a share of booked visits. Contracts as a share of completed visits. Average contract value split between repair and replacement. And all of it split at minimum between inside the Perimeter and outside it. Review them weekly. Almost every decision worth making becomes visible in that set.
