Every call should end with three facts and a date on the calendar
An intake call that produces a name and a callback promise has produced nothing.
Write down the address, the covering and whether water is coming in right now. Those three facts decide who goes, what they bring and how urgent the visit is.
Then get a date. Not a promise to call back with a date. A specific window, confirmed by text before the call ends, because a scheduled appointment survives a competitor's phone call and a vague promise does not.
Give whoever answers a script with the questions in order and the honest answers to the two objections that come up every time: when can you come, and what will it cost to look. If the answer to the second one is nothing, say so early.
Assign the call. In small companies the phone belongs to whoever is nearest, which means it belongs to nobody. One person owns intake during working hours and a monitored service owns it outside them.
A rained out inspection is not rescheduled until somebody owns the rebook
Weather cancels appointments constantly here, and the rebooking step is where a surprising number of jobs quietly end.
When rain moves an appointment, the customer hears silence and assumes they are back at the start. Meanwhile the estimator's day fills with the jobs that did not get canceled.
Make the rule explicit: any appointment canceled for weather gets a new date offered the same day, in writing, before the office closes. Not a callback later in the week.
Track how many rained out appointments were rebooked within forty-eight hours. The gap between that number and the number canceled is a list of jobs you paid to generate and then handed to somebody else.
Use the delay rather than apologizing for it. A short message explaining what the crew is doing while it rains, and what you will check on the roof when the weather clears, keeps the appointment alive.
In the Garden District the wait for approval is where a signed job goes quiet
Work on a visible exterior in a historic district can sit for weeks between the handshake and the start date, and silence during that stretch produces cancellations.
Set the expectation at the proposal stage rather than after the customer starts wondering. Explain the sequence in general terms and say plainly that timing is not fully in your control, and that it is worth confirming details with the relevant commission or the city.
Build a scheduled sequence of short updates for the waiting period. What was submitted, what you are waiting on, what happens when it clears. Two sentences at a set interval beats one long email nobody expects.
Give the customer something to do. Color and material selections, scheduling preferences, a walkthrough of where the dumpster and ground protection will go. A customer participating in a project does not shop it.
Track approval waits separately in your pipeline so they do not look like stalled sales. A job waiting on a commission and a job waiting on a decision are different problems, and blending them makes your close rate unreadable.
Gutters and where the water lands belong in the proposal, not in a change order
In a city built around drainage, the conversation about water leaving the roof is part of the sale rather than an add-on that surprises somebody in week two.
Homeowners here already think about water. A high water table, low elevation and constant reliance on drainage make gutter capacity and downspout discharge ordinary household topics, not upsells nobody asked for.
Put the drainage discussion in the proposal, itemized and optional if you prefer, with the reasoning written out. A customer who declines an item they understand is far less likely to argue about it later than one who first hears about it from a crew leader.
Do the same with the unknowns. Decking condition under the covering is the classic mid-job phone call, so explain how you handle it, at what rate, and what triggers a call before anyone starts.
Stay in your lane on anything structural. If a roof plane is moving or a question turns to the foundation, say so honestly and recommend a licensed engineer rather than offering an opinion. Referring the question out builds more trust than answering it would.
Give every lost roof a reason code and the pattern shows up in a month
You cannot fix a leak in the pipeline you cannot name, and free-text CRM notes never aggregate into anything.
Use a short fixed list of reasons and require one on every closed lost job. Price, timing, went with another contractor, never reached them, out of area, waiting on a decision, not our type of work.
Add the parish and the job type to the record. Losing Orleans repairs for a different reason than St. Tammany replacements is the kind of pattern that stays invisible in a monthly lead total.
Review the list once a month with the office and one estimator in the room. Twenty coded losses tell you more about what to fix than a dashboard averaging six months of everything.
Count what actually matters at each stage: inquiries received, inspections completed, proposals delivered, contracts signed. The steepest drop between two of those numbers is your project for the quarter, and it is usually not the one everybody assumed.
Put a stopwatch on the pipeline, because the traffic will never settle a split test
A single metro rarely produces enough volume to prove a small change, so measure elapsed time between steps and attack the slowest one.
Record four timestamps: inquiry received, first human contact, inspection completed, proposal delivered. Nothing more complicated than that, and any CRM can hold them.
Look at the medians after a month. Time to first contact is usually the worst number in a roofing company, and it is the one most directly connected to whether a customer keeps calling other contractors.
When you do want to change something, change a whole path rather than a detail. A new intake script, a rewritten proposal, a follow-up sequence with real content in it. Effects large enough to see without statistics are the only ones this volume can confirm.
Supplement the numbers with recordings and short calls to people who did not hire you. Ten honest conversations with lost prospects in Metairie or Algiers will surface things no report contains, and they cost an afternoon.
Questions we actually get
- What is the single highest value fix for most roofing companies here?
- Answering the phone and booking a specific appointment on that first call. Nearly every other improvement depends on it, and most companies overestimate how well they do it. Pull a week of call records and count how many ended with a confirmed date rather than a promise to follow up.
- How many follow-ups should a proposal get before we stop?
- More than one, spaced out, and each carrying something useful rather than a request for a decision. A photo from the inspection, an explanation of a line item, an answer to a question they raised. Stop when they tell you to stop or when the reason code says the job went elsewhere.
- Should we run A/B tests on our website?
- For button colors and headline variations, one metro almost never produces the volume to reach a trustworthy result, and a test that runs for months tells you about a season instead of a change. Test whole processes: the intake script, the proposal format, the follow-up sequence. Judge them on signed roofs over a comparable period.
- A customer thinks the roof is causing a problem that might be the house itself. How should the office handle it?
- Answer honestly and hand it off. Describe what your crew can see from the roof, avoid diagnosing anything structural, and recommend a licensed engineer for questions about movement or foundations. Being the company that says it does not know tends to be remembered well.
- How do we track whether follow-up is actually happening?
- Make it a field rather than a habit. Every open proposal carries a next action and a date, and the weekly meeting reviews anything past due. If your CRM cannot produce a list of proposals with no scheduled next step, that report is the first thing worth building.
