The leak call is qualified by address before it is qualified by budget
Jurisdiction is the first question in this metro and asking it late wastes an estimator's whole afternoon.
The District, suburban Maryland and Northern Virginia license and permit separately, and specifics are worth confirming with each jurisdiction. Whatever your coverage is, the person answering the phone needs to establish which side of a line the property sits on inside the first minute.
Write it into the script as a natural question rather than a gate. Asking for the property address early is normal, useful and lets you say quickly whether this is a job you can take. Everybody prefers a fast no to a slow one.
Then route by covering. A flat rowhouse roof and a pitched shingle roof often need different estimators, different equipment and different scheduling, and sorting that on the call prevents the survey where the wrong person shows up.
Track declines with a reason code. If a meaningful share of inquiries are being turned away for jurisdiction, that is a targeting and website problem showing up at the phone, and it will keep costing you until the source is fixed.
After a thaw, the first roofer who answers gets the rowhouse
Response time decides more of these jobs than anything on the proposal, and the pressure arrives in bursts tied to weather.
Freeze and thaw cycles produce clustered leak calls. Everybody in the trade gets busy in the same week, and homeowners work down a list until somebody picks up. Being second on the phone is usually being second everywhere afterward.
Plan coverage for those weeks specifically. Whoever answers should be able to book a survey directly rather than take a message, and there should be a named fallback when the volume spikes. An answering service that only takes names converts far worse than one that can put a time in the calendar.
Set an internal standard for first contact on web inquiries and measure against it. Most roofers assume they are fast. Pull the timestamps for a month and look, because the gap between assumption and reality here is usually where the money went.
After hours matters. Ceiling stains get noticed in the evening. Decide whether you are answering at night, and if you are not, make sure the automatic reply says exactly when a human will call and then keep that promise.
A survey appointment in Georgetown needs a parking plan, or it does not happen
Failed and late surveys are a scheduling problem dressed up as a sales problem.
Booking a rowhouse survey without asking about parking and access is how an estimator ends up circling Capitol Hill for twenty minutes. Ask on the call: driveway, street with permit parking, alley behind, or rear access through a neighbor.
Cluster appointments by jurisdiction and by day. Sending one estimator from Reston to the District and back for a single survey costs more in windshield time than most companies ever put in the margin calculation. Group Northern Virginia days and District days.
Confirm the appointment twice, and confirm what the owner needs to do. Attic access, a key, someone home, a gate. Small confirmations sharply reduce the survey that arrives and cannot get on the roof.
Count completed surveys, not booked ones. Booked appointments flatter everybody. The completed survey is the first point where the pipeline is real, and it is the number the whole intake process should be built to increase.
Federal and contractor buyers read a proposal the way they read a solicitation
This audience rewards scope, credentials and pricing logic, and punishes closing pressure.
Many households here are used to procurement style diligence. They compare documents rather than personalities, they want to see what is included and excluded, and they are entirely comfortable asking for insurance certificates and license information.
Structure the proposal accordingly. Scope, assembly, what is being removed, what is being installed, what happens if the deck is bad underneath, warranty terms, payment schedule. A clear exclusions list builds more trust than it costs.
Drop the expiring discount. With this buyer, artificial urgency reads as a reason for caution, and it will get discussed with a spouse who works in contracting. Give a real reason for timing instead, such as crew availability or the weather window before a freeze.
Follow up on a schedule and say what the schedule is. Two or three planned contacts over a defined period, each adding something useful, outperform the single hopeful call most roofers make. Say nothing about insurance outcomes on a specific claim, and point coverage questions to their insurer or their own counsel.
Tag every inquiry with its jurisdiction or your reporting is fiction
A blended pipeline across three legal markets averages away every difference that matters.
Add a jurisdiction field to the CRM record and make it required. District, Montgomery, Prince George's, Arlington, Alexandria, Fairfax, and whatever else you serve. Nothing else you do to reporting will produce as much clarity for as little effort.
Then look at the funnel by jurisdiction: inquiry, qualified, survey completed, proposal sent, job booked. Close rates and cost to serve usually differ enough between a District rowhouse and a Fairfax County subdivision that seeing them merged has been actively misleading you.
Add a second field for covering. Low slope, shingle, slate, metal. The combination of jurisdiction and covering tells you which marketing to buy more of far better than any lead count.
Review call recordings monthly with the intake person, not at them. Listen for the point where the caller hesitates. In this market it is often a coverage question or an access question, and both are fixable in a sentence.
Test the intake script and the follow-up, not the shade of the button
One metro rarely generates enough traffic to settle a small design difference, so spend the effort where the effect is large.
Split testing needs volume that most single market roofing companies do not have. Waiting months for a result on a headline variation is not a program, it is a delay, and the answer often arrives too late to be worth acting on.
Test big, visible changes instead. Adding a jurisdiction selector to the form. Moving coverage above the fold. Adding an access question. Changing who answers the phone during thaw weeks. Effects that size show up in the numbers without statistical acrobatics.
Change one thing at a time and give it a defined period, then write down what happened before you change the next. The discipline matters more than the sophistication of the method.
Talk to people. Five conversations with recent customers about what nearly stopped them from calling will produce more usable direction than a quarter of underpowered experiments. Ask what they searched, who else they called, and what made them pick up the phone.
Questions we actually get
- What is the single biggest cause of lost roofing inquiries here?
- The two usual candidates are slow first response during freeze and thaw weeks, and surveys that never complete because access was not established when the appointment was booked. Both are process fixes rather than marketing fixes, and both are measurable from data you already have in the phone system and the calendar.
- Should we qualify by jurisdiction before we take contact details?
- Ask for the property address early, but take the details either way. A caller you cannot serve is still worth a referral and a polite ending, and a form that gates on jurisdiction before capturing anything will lose people who mistype a ZIP. Capture first, route second.
- How many inquiries do we need before split testing is worthwhile?
- There is no clean threshold, and any specific number quoted without seeing your data is invented. The practical test is whether a change would have to be large to be visible. If yes, make large changes and read them carefully. If your volume genuinely supports fine grained tests, your analytics will tell you so.
- Is call recording worth setting up?
- Generally yes, and it is usually the fastest source of insight available. Recordings show where callers hesitate and which question the intake person answers badly. Confirm your recording and consent obligations for each jurisdiction you operate in with your own counsel, because the rules are not identical across the District, Maryland and Virginia.
- Our close rate drops on larger proposals. What should we look at first?
- Look at the time between the survey and the proposal, and at whether the proposal explains the assembly and the exclusions. In this market a document that reads like a scope of work holds up better in a household comparison than one that reads like a sales quote, and delays after a survey are where interest cools fastest.
