How to Fill a New Roofing Crew's Schedule From Scratch

You signed the crew. Maybe you poached a sharp foreman, hired three guys he vouched for, leased a dump trailer, and wrote the first week's payroll check before a single shingle hit a tear-off. Now it's Sunday night and the board for that crew has one job on it. Monday afternoon they'll be standing in the yard looking at you.

That gap, between the day labor starts costing money and the day the schedule is reliably full, is where new roofing operations bleed out. It is rarely a quality problem or a hustle problem. It is a math problem and a sequencing problem, and almost nobody writes it down. They wing it, burn cash for ninety days, and either get lucky with a storm or run the owner's credit line to zero.

What follows is the version I wish someone had handed me: the capacity math, the channels that actually fill a board fast versus the ones that pay off in month six, the canvassing and routing mechanics, the scripts, and a day-by-day ramp plan. It assumes you are a real contractor who knows how to roof and now has to keep people fed. No theory you can't run on a Tuesday.

Start with the number: what "full" actually means

Before you generate a single lead you need to know the target. "Keep the crew busy" is not a target. A target is a number of booked, signed, materials-ordered jobs per week that keeps that specific crew producing without gaps.

Work it backwards from production capacity.

A standard residential tear-off-and-reroof crew of 4 to 6 people, running asphalt shingles, completes roughly one average-size roof per day in good weather. Average residential roof in most US markets runs 22 to 30 squares (a square = 100 sq ft). Call it a 25-square house as your planning unit. So a single crew, five working days, no rain, no callbacks, is about 5 roofs or ~125 squares per week of installed capacity.

Reality knocks that down. Build in:

• Weather loss. Even in a dry market you lose 15 to 25 percent of days to rain, wind, or wet decking. Storm markets can be worse.
• Travel and setup. Spread-out jobs cost you a half day moving and staging.
• Punch-out and callbacks. A crew that never goes back for a missed pipe boot doesn't exist.

So your effective booked capacity per crew is closer to 3.5 to 4 completed jobs per week, not 5. That is the number you have to feed.

Now the pipeline rule almost everyone gets wrong: you do not need 4 leads a week to install 4 jobs a week. You need enough leads to net 4 signed contracts with the deposit and materials moving, after every fall-off. Let's build that.

The pipeline math, worked

Here is a realistic retail (non-storm) funnel for a competent salesperson knocking and running set appointments:

Run it backwards. To install 4, you need ~4.5 signed (allowing 10% fall-off). To sign 4.5 at a 35% close, you present ~13 estimates. To present 13, you sit ~16 appointments. To set 16 appointments at 60%, you need ~27 qualified leads per week, per crew.

That is the headline number nobody tells the new owner: roughly 25 to 30 genuine, qualified opportunities a week to keep one crew full at retail close rates. If your close rate is better (storm work, strong brand, insurance-driven), the lead requirement drops fast. At a 55% estimate-to-sign close, the same crew fills on ~16 to 18 leads a week.

Make your own version of this table in a spreadsheet on day one. Track the real conversion at each stage for your people. The first time you measure it, your appointment-set rate will be lower than 60% and your close will be lower than you think. That is fine. You can't fix what you haven't counted.

Put a dollar figure on the empty Monday

The abstract version of "keep the crew busy" doesn't motivate anyone. The dollar version does. So price the idle day before you do anything else, because that number is what justifies your marketing spend and your sense of urgency.

Take a 5-person crew. Loaded labor (wages plus payroll tax, workers' comp, and the slice of overhead that crew carries) runs, conservatively, $1,500 to $2,500 a day depending on your market and pay structure. A crew that sits one day a week for the first six weeks burned somewhere between $9,000 and $15,000 of payroll for zero production. That is real money you already committed to. It does not come back.

Now flip it. That same crew, producing at 3.5 to 4 average jobs a week, is generating revenue. If your average residential job is, say, $12,000 to $16,000 and you run a healthy gross margin, a single full week of that crew produces enough gross profit to fund a month of aggressive canvassing and mail. The lesson is blunt: lead-gen spend that fills the board is almost never your most expensive line item, the idle crew is. New owners get this backwards, they pinch on marketing to "save money" while paying a full crew to stand in the yard. Spend the marketing money. The math is on its side.

Cost per acquisition, and the only marketing number that matters

The number you steer by is cost per acquired job (CPA), total marketing-and-sales cost in a period divided by signed jobs in that period. Early on it'll be ugly because the slow clock hasn't kicked in. That's expected. What you're watching is the trend. A healthy ramp shows CPA falling month over month as reviews, referrals, and repeat-neighborhood presence start producing leads you didn't pay cash for.

A simple way to keep it honest: tag every signed job with its source (canvass, mail, referral, GBP, paid lead, past customer). Once a month, total spend and signed jobs by source and compute CPA per channel. You'll usually find your warm and canvassed jobs cost a fraction of your paid-lead jobs, which tells you exactly where to push hours and where to pull back dollars. Don't guess which channel works. Tag, total, and let the per-channel CPA decide your next month's budget.

The capacity-vs-pipeline balance

There are two failure modes and they are mirror images:

1. Under-fed crew. Pipeline is dry, crew sits, payroll burns, foreman gets nervous and starts shopping for another company. This kills new crews.
2. Over-sold crew. Sales got hot, you booked six weeks out, and now homeowners who signed in week one are calling in week four asking where you are. Cancellations climb, reviews sour, deposits get refunded.

The healthy state is a 2-to-3 week backlog for one crew. Enough that a couple of rain days don't empty the board, not so much that customers walk. Watch that backlog number like a fuel gauge. When it drops under one week, you spike lead-gen spend and canvassing hours. When it climbs past three to four weeks, you either slow marketing or, better, you're ready to staff a second crew.

The two clocks: fast-fill channels vs. compounding channels

Every lead channel sits somewhere on a speed-versus-cost-versus-durability spectrum. The mistake new owners make is pouring everything into channels that pay off in month six when they need bodies producing this month.

Think in two clocks.

Fast clock (fills the board in days to weeks):

• Door knocking / canvassing
• Your existing relationships and database (past customers, friends-and-family, suppliers, real estate and property managers)
• Targeted direct mail and door hangers to qualified streets
• Paid lead aggregators (use carefully, see below)
• Paid search for high-intent terms ("roof repair near me")
• Storm response when a real event hits your area

Slow clock (compounds for months/years, low marginal cost later):

• Google Business Profile + reviews
• Organic SEO / your website ranking
• Referral program that actually pays out
• Branded trucks, yard signs, jobsite presence
• Social proof / community reputation

For a brand-new crew you run BOTH at once but you fund the fast clock and you plant the slow clock. The fast clock keeps the lights on for the first 90 days. The slow clock is what makes lead-gen cheap by month six so you're not forever renting leads.

The rest of this is organized around exactly that: first the fast-fill channels in the order you should deploy them, then the routing and scheduling mechanics, then the slow-clock plant, then the 90-day plan.

Channel 1: Your own database and relationships (day one, free)

Before you spend a dollar on marketing, mine what you already have. Even a brand-new company has a network, and a new crew inside an existing company has the whole house list.

Past customers and warm contacts

If you have any completed jobs, those homeowners are your single highest-converting source. They know your work. Their neighbors saw your sign. Pull every past customer and do three things:

1. Maintenance/inspection offer. "We're in your area next week doing free roof check-ups for past customers, want us to take a look?" This isn't a gimmick if you actually inspect and report honestly. Older customers' roofs are aging; some genuinely need work.
2. Referral ask, specific. Not "send me referrals." Instead: "Do you know anyone on your street or at work whose roof took a beating, or who's mentioned theirs is getting old? I'll take great care of them."
3. Review ask if you never collected one. Feeds the slow clock.

No past customers at all? Then your warm list is suppliers, your subs, your insurance agent contacts, real estate agents, property managers, your kid's coach, your church, the trades you know. Tell every one of them, specifically, that you have a crew with open capacity right now and you're looking for roofs. People help when you make the ask concrete and time-bound.

Property managers and small commercial

Property managers control dozens of roofs and they hate emergencies. A new crew with open Tuesday-Wednesday capacity is exactly what a PM wants for their flat-roof leaks and small re-roofs. One signed PM relationship can quietly keep a crew partly fed for a year. Cold-call the property management companies in your county, lead with availability and responsiveness, not price.

Channel 2: Canvassing — the fastest board-filler there is

Door knocking is unglamorous and it is the single most reliable way to put jobs on a new crew's board this week without a marketing budget. It converts attention into appointments with zero ad spend and a same-day feedback loop. If you only master one fast-clock channel, master this.

But random knocking is a grind with terrible yield. The difference between a canvasser who nets two appointments a day and one who nets eight is almost entirely where they knock, not how charming they are. Targeting is the whole game.

Where to knock (the targeting decision)

You want streets where the density of roofs that are actually due is high. Two independent signals stack here:

• Roof age. A subdivision built in 1999 to 2003 with original 3-tab or first-gen architectural shingle is a wall of roofs hitting end of life at the same time. Whole streets aging out together.
• Storm exposure. Neighborhoods that took a real hail or high-wind event have a cluster of roofs with legitimate impact damage the homeowner may not have looked at.

When both line up, the same tract, same build era, recent storm, your knock-to-appointment rate can double or triple versus knocking blind. You are talking to people whose roofs are genuinely near the end, instead of arguing with someone who reshingled two years ago.

The old way to find these streets is local knowledge and windshield time: you drive the tract, eyeball granule loss and curling, note the build year from county records, remember which way the last storm tracked. That works and you should still do it. The slow part is doing it at scale across a whole metro, street by street.

This is the natural place for data to do the legwork. Tools like RoofPredict score addresses by an estimated roof-age range read from aerial imagery and overlay storm physics modeled per individual roof, then rank streets and lists so your canvassers spend the day in front of the doors most likely to be due, instead of burning shoe leather on a tract that got reroofed last spring. Be clear-eyed about what that buys you and what it doesn't: roof age comes back as a range, not a build certificate, and storm modeling gives you odds a given roof was affected, not proof of damage. It points the truck at the right blocks; your foreman's eyes and a ladder still decide whether a specific roof has a real, documentable problem. Used that way, it turns canvassing from a blind grind into a routed, prioritized day, which matters enormously when every crew-day you can't fill is real money out the door.

Whether you buy data or build the list by hand, the targeting rule is the same: knock streets where roofs are due, not random streets.

The canvassing daily structure

A productive canvasser's day, whether it's you, your sales hire, or a small canvassing team:

1. Pre-built route, not freestyle. Each rep gets a defined cluster of streets for the day, ideally pre-scored or pre-driven. No deciding where to go at 9am.
2. Knock the productive hours. Best residential windows are roughly 10am-1pm (retirees, shift workers, work-from-home) and 4pm-7:30pm (everyone home). Midday is for following up and dropping door hangers on no-answers.
3. Density over distance. Forty doors on three connected streets beats forty doors scattered across town. Travel time is dead time.
4. Track every door. Knocked, not-home, not-interested, appointment-set, inspected. You cannot manage canvassing you don't measure.

A worked canvassing day, with the numbers

Let's make the yield concrete so you can set rep expectations and staff correctly. A focused canvasser working productive hours can physically knock 40 to 70 doors in a day. Not all answer. Here's a realistic blind-knock day versus a targeted-knock day on the same effort:

Same legs, same hours, two-to-three times the productive output, purely because the targeted day put the rep in front of roofs that were actually near end of life or genuinely storm-exposed. Over a five-day week that's the difference between a rep netting ~5 opportunities and one netting ~15. One rep on good lists can come close to feeding a whole crew. That gap is exactly why the targeting work in Week 0 matters more than the pep talk.

Staffing and paying canvassers

For a new operation you have three options, in rough order of how most owners start:

1. You and your foreman knock. Free, highest-skill, but it pulls your best people off production and off running the business. Fine for the first few weeks; not scalable.
2. A dedicated canvasser or two. Often paid a modest base plus a per-set-appointment or per-signed-job bonus. Structure pay so the rep is rewarded for qualified appointments that actually sit, rather than raw doors knocked, or you'll get inflated activity and no-show appointments.
3. A small canvassing crew with a closer behind them. Canvassers set inspections; a stronger salesperson does the roof climb, the documentation, and the presentation. This separates the skill of getting permission from the skill of closing, and it scales.

Whatever you choose, manage canvassing with a daily number, not vibes. Each rep reports doors knocked, contacts made, inspections set, and appointments sat. Two weeks of that data tells you who can canvass and who can't, and it tells you whether your lists are the problem versus your people. If a good rep is getting weak yield, suspect the list before you blame the rep.

A canvassing script that holds the line (and stays legal)

The goal at the door is not to sell a roof. It is to earn permission to get on the roof and document it. Here is a frame that works and stays on the right side of the law:

"Hi, I'm [Name] with [Company]. We're doing roof inspections on [Street] this week. A lot of the roofs in here are the same age and we've been finding wear and some storm damage. It's a free inspection and I'll show you photos of exactly what your roof looks like, good or bad. Takes about twenty minutes, want me to take a look while I'm here?"

What that does: time-bound, neighborhood-specific, free, low-commitment, and it promises documentation (photos), not a payout. That last part matters.

The legal line you must coach into every rep: when storm or insurance comes up, your people inspect, photograph, and document. They do not promise the homeowner the insurance company will approve a claim, do not promise "you'll get a free roof," do not say the deductible will be waived or absorbed, and do not offer to handle, negotiate, or "fight" the claim with the carrier on the homeowner's behalf. That last set of moves is unlicensed public adjusting in most states and it can end your business.

The compliant pitch is simple and honest: we document the roof thoroughly, we write you an accurate repair estimate, and we hand it to you. You file with your insurer, and your insurer decides what's covered. You're the documentation and repair expert, never the claims handler. Train this on day one, because a green canvasser trying to close will reach for "I'll get you a free roof" by instinct, and that one sentence is a liability.

Channel 3: Targeted direct mail and door hangers

Mail is the canvassing cousin: same targeting logic, no person required, scales while your reps sleep. It is slower than knocking but it stacks. The trick, again, is the list.

Blasting an entire ZIP at $0.50 to $1.00 a piece is how new owners waste money. A targeted drop to streets that score high for roof age and storm exposure, the same lists you'd canvass, gets several multiples of the response of a blind blast because most of those homeowners actually have a due roof.

Practical mail playbook:

• Match mail to your knock routes. Mail a street a few days before you knock it, then the door hanger and the postcard reinforce each other. Recognition lifts the knock.
• Door hangers on every no-answer. Your canvassers should leaf every not-home door. It's a second impression at zero added travel cost.
• One clear offer, one phone number, a QR to a real booking page. Free inspection, local company, fast scheduling. Don't get cute.
• Expect 0.5% to 2% response on a targeted list, far less on a blind one. Budget accordingly: a 1,000-piece targeted drop might net 5 to 20 calls and a handful of jobs. Do it weekly, in rhythm with your routes.

Channel 4: Paid leads and paid search (fast, but priced like it)

When the board is empty today and you have some cash, paid channels buy speed. Just understand exactly what you're buying.

Lead aggregators (the rented-lead trap)

Services that sell you "roofing leads" can fill appointments fast. The catch:

• The lead is usually sold to 3 to 5 contractors at once, so you're racing four other trucks and the homeowner is price-shopping.
• Quality is uneven, you'll pay for tire-kickers, wrong numbers, and renters.
• Close rates run low (often 10-20%), so the effective cost per signed job can be brutal.

Use them as a stopgap to bridge a gap, not a foundation. The rule: call within five minutes, every time. Shared leads go to whoever calls first. If you can't call back in minutes, don't buy shared leads at all.

Paid search (Google Ads / LSAs)

Google's Local Services Ads (the "Google Guaranteed" units at the top) are often the best paid spend for a roofer because you pay per lead, the leads are local and high-intent, and the badge builds trust. You'll need to pass Google's license/insurance screening. Standard search ads work too but need tighter management and a real landing page.

For a new crew, paid search is a fine fast-clock supplement, but treat it as rented traffic. The moment your organic presence and reviews mature (slow clock), your blended cost per job drops and you lean off paid.

Channel 5: Storm response (when it's real)

If a legitimate hail or wind event hits your service area, that is the single fastest board-filler that exists, a whole region of damaged roofs at once. Be ready to surge canvassers into the impacted footprint within 24 to 48 hours.

Two cautions:

1. Verify the event is real and target the actual swath. Storms are narrow and patchy. Knocking three miles outside the hail core wastes the surge. Storm-track and per-roof storm-physics data (NOAA/NWS reports, plus per-address modeling like RoofPredict's) keeps your crews inside the real damage footprint instead of guessing from a news map.
2. Stay compliant under pressure. Storm chasers cut corners and that's when the "free roof" and "we'll handle your claim" promises come out. Same rules apply, harder: document, estimate, hand off. Let the homeowner file and the insurer decide. Reputable storm work is just retail roofing with a tighter clock and better-qualified roofs.

The inspection-to-estimate workflow that converts (and keeps you compliant)

Filling a board isn't just generating leads, it's converting the inspections those leads produce into signed jobs without crossing any legal line. The roof climb is where the sale is won or lost, and where new reps most often say something that creates liability. Build a tight, repeatable workflow and train it hard.

What a documentation-grade inspection looks like

Whether it's a storm roof or an aging retail roof, the inspection produces a record, not a verbal opinion. The homeowner should walk away with photos and a written scope they can hold in their hand. A solid inspection captures:

• Overview shots of every roof plane and elevation, plus a wide shot of the house with address visible.
• Close-ups of every defect with something for scale (a chalk circle, a coin, your hand). Granule loss, mat exposure, bruising/soft spots from hail, creased or missing shingles from wind, cracked or worn boots, failed flashing, nail pops, deck issues at eaves.
• The accessories, gutters, downspouts, soft metals, screens, AC fins, often show collateral storm impact that corroborates roof damage. Photograph them.
• Slope and direction context, note which elevations face the storm's approach. This matters for an honest read of what's wear versus what's impact.
• Measurements, either a tape-and-pitch read or an aerial measurement report, so the estimate that follows is accurate, not a finger in the air.

Date-stamp and address-tag the photos. A clean, organized photo set is the single most persuasive thing you can put in front of a homeowner, and it's the thing that makes your repair estimate credible.

Writing the estimate the right way

From the documentation you produce an accurate repair estimate for your scope of work, aligned to standard pricing practice (an Xactimate-aligned line-item estimate is the norm for storm work, and it reads as professional even on a retail job). The estimate describes what you will do to put the roof right and what it costs. That's it. You're stating facts about your own scope.

Then you hand it to the homeowner. On a storm job, the homeowner uses your documentation and estimate when they file with their insurer, and the insurer decides what is covered. You are the documentation-and-repair expert. You are not the claims handler.

The do-not-say list, taught as compliance, not as a wink

This is worth putting on a laminated card in every truck because the violations are sales reflexes, not malice. Your reps must never:

• Promise the insurer will approve the claim or that it "definitely qualifies."
• Promise a "free roof."
• Promise the deductible will be waived, absorbed, eaten, or "taken care of" (paying or rebating a customer's deductible is insurance fraud in most states and explicitly illegal in many).
• Offer to negotiate, adjust, handle, or fight the claim with the carrier on the homeowner's behalf, that's unlicensed public adjusting, which is licensed and regulated state by state.
• Interpret the policy or tell the homeowner what their coverage means.

What they can and should say: we'll document the roof thoroughly, we'll write you an accurate estimate to repair it, and we'll hand it to you so you have a complete, professional record. You file the claim with your insurer, your insurer's adjuster reviews it, and they decide coverage. If they approve repairs, we're ready to do excellent work. This frame captures every bit of the homeowner's real intent (they want help dealing with a damaged roof) while keeping you firmly on the documentation-and-repair side of the line. It also, frankly, closes better, because homeowners trust the contractor who explains the process honestly over the one promising a free roof they can sense is too good to be true.

Now the other half: routing, scheduling, and protecting the crew's time

Filling the board is only half the job. A full board that's badly sequenced still leaves the crew idle, because of a rain delay with no backfill, a job that wasn't material-ready, an inspection two counties away that ate a half day. Lead-gen gets the attention; scheduling discipline is what actually keeps the crew producing.

Job-readiness gates: nothing hits the board until it's ready

A job is not "scheduled" just because it's signed. It's schedulable when:

• Contract signed, deposit collected (or financing approved).
• Materials measured, ordered, and delivery date confirmed.
• Permit pulled if your jurisdiction requires it (build the permit lead-time into the schedule, some cities are days, some are weeks).
• HOA/color approvals cleared if applicable.
• Crew access and dumpster/trailer logistics sorted.

Make a literal checklist. A job that fails any gate goes to a "pending" column, not the production board. The classic new-crew killer is a foreman showing up to a driveway with no delivered materials.

Cluster jobs geographically

When you schedule the week, batch by area. A crew doing three jobs in the same subdivision on consecutive days saves hours of windshield time and lets one materials delivery serve multiple roofs. This is where targeted canvassing pays a second dividend: if you knocked one tract hard, you signed clustered jobs, which schedule tight. Scattered sales mean a scattered, inefficient board.

Always carry a backfill list (the rain plan)

Keep a running list of jobs that are ready to go but flexible on date so that when Tuesday rains out, you slide Thursday's covered-porch tear-off up, or you send the crew to that detached-garage reroof that doesn't care about a little wind. Repairs, smaller jobs, and partials make great backfill. A crew should almost never sit because of weather if your pending column has ready work in it.

Buffer days and the over-promise trap

Don't book five installs into five days. Book four and leave a floating buffer day for weather slip, callbacks, and the job that runs long. A schedule with zero slack turns one rainy Monday into a cascading three-week delay and a wall of angry customers. Under-promise the start date by a few days; finishing early is a five-star review, finishing late is a refund request.

Capacity triggers for crew #2

Watch two numbers together: backlog length and lead surplus. When your backlog sits consistently at 3-plus weeks AND you're turning away or slow-rolling leads you're paying to generate, that's the signal to staff a second crew, not before. Adding a crew before the pipeline can feed it just doubles your idle payroll. Adding it too late means you're refunding deposits and burning the reviews you worked for. The data should drive the timing.

The schedule board itself: keep it dumb and visible

You do not need expensive software to run one crew. You need a board everyone can see and trust. A shared calendar or a simple roofing CRM with a job-status pipeline works. What matters is the columns, not the brand:

• Leads / inspections set (not yet sold)
• Sold, pending (signed but failing a readiness gate, awaiting materials, permit, color, financing)
• Ready to schedule (all gates passed, can hit a date)
• Scheduled (on a specific day, with crew and materials confirmed)
• In progress
• Punch-out / callback
• Closed / invoiced

The rule that keeps this honest: a job moves from "sold, pending" to "ready to schedule" only when every readiness gate is checked, and only "ready" jobs go on a date. This one discipline prevents the most common new-crew disaster, a foreman in a driveway with no materials, because un-ready work physically can't reach the production board.

Your weekly operating rhythm

The owners who keep crews full run a tight weekly cadence instead of reacting to whatever catches fire. A workable rhythm:

• Friday afternoon: lock next week's crew schedule from the "ready" column, clustered by area, with a buffer day and the backfill list confirmed. Order any materials not yet ordered.
• Friday/weekend: review the gauges, backlog length, this week's signed jobs, conversion by stage, CPA by channel. Decide whether to spike or ease canvassing and mail next week.
• Daily (AM): quick crew check, materials confirmed for today, weather check, backfill ready if rain.
• Daily (PM): canvassers report their numbers; new solds get gated and slotted; review requests sent on completed jobs.

This cadence is what turns the playbook from a one-time scramble into a machine. The gauges tell you when to feed the pipeline; the readiness gates keep the board clean; the buffer and backfill keep weather from idling the crew. Run it every week and the empty Monday stops recurring.

Enriching your own list: turn a CRM into a target map

A point most new owners miss: you probably already have a list, a CRM of past customers and old quotes, a mailing list, a territory you've worked, and it's sitting there flat, every address treated equally. It isn't equal. Some of those addresses are roofs that have aged five more years since you last touched them. Some sit inside a storm footprint you didn't track.

Enriching that list, appending a roof-age range and a per-roof storm-exposure signal to addresses you already own, turns a flat database into a ranked target map. Instead of mailing your whole old-quote list evenly, you mail and re-knock the slice where roofs are now due. This is where data like RoofPredict's earns its keep on the retention and reactivation side, beyond net-new canvassing: it scores the addresses in your own CRM or mailing list by roof-age range and modeled storm physics so you re-engage the homeowners most likely to need work now. Same honest limits apply, age is a range, storm is odds, the ladder makes the final call, but it means a crew you stand up doesn't start from a cold blank list. It starts from a ranked one, including the prospects you'd already paid to acquire once.

Planting the slow clock while the fast clock runs

Everything above fills the board now. Do these in parallel so that by month four your cost per job falls and you stop renting every lead.

Google Business Profile and reviews (the highest-ROI slow play)

For a local roofer, a fully built Google Business Profile with a steady stream of recent reviews is worth more than almost any paid channel over time. Set it up week one. Then make review collection a process, not a hope:

• Ask at the moment of peak happiness, the walkthrough when the new roof looks great and the yard is clean.
• Hand them a card with a QR straight to your review link, or text the link before you leave the driveway.
• Aim for a review on a meaningful share of completed jobs. Even one a week compounds. A roofer with 80 recent 5-star reviews out-converts a stranger with 4, on the same street, every time.

Yard signs, branded trucks, jobsite presence

Every job is a billboard. Put a sign in the yard the day you start and leave it a week after. Wrap or letter the trucks. When you've clustered jobs in a subdivision (because you targeted it), the neighbors see your name five times in two weeks, and the next knock is warm.

Referral program that actually pays

A referral that closes is your cheapest, highest-closing lead. Make the offer concrete and pay it fast, a real check or gift card the week the referred job installs. Vague "we appreciate referrals" does nothing. "$200 the week your neighbor's roof is done" gets you neighbors.

A real website and basic SEO

You need a fast, mobile, trustworthy site with your service area, real photos of your work, your reviews, and an easy way to book an inspection. It's where every door hanger QR, every Google click, and every referral check goes to confirm you're real. SEO ranking takes months, that's the slow clock, but the site has to exist on day one to catch the traffic the fast clock drives.

Putting it together: a 90-day ramp plan for one new crew

Here's the whole thing sequenced. Adjust the numbers to your market and your close rate, but follow the order.

Week 0 (before the crew starts)

• Build the pipeline math sheet: your real target is ~25-30 qualified leads/week to fill one crew at retail close (fewer if you close higher).
• Stand up Google Business Profile, a basic booking-capable website, and your review-collection process.
• Mine your database and warm network, line up the first week's inspections from past customers, friends, suppliers, property managers.
• Pull or build your first targeted canvassing/mail lists: streets scored by roof-age range and storm exposure. This is where age + per-roof storm data earns its keep, point the truck at due roofs from day one.
• Order signs, hangers, and letter the trucks.
• Train every rep on the at-the-door script and the legal do-not-say list (no promised approvals, no "free roof," no deductible promises, no claim handling).

Weeks 1-3 (fill the board, all gas on the fast clock)

• Canvass hard on pre-built routes, density over distance, productive hours, every door tracked.
• Mail the streets a few days ahead of the knock; hangers on every no-answer.
• Work the warm list for the first easy signs.
• If cash allows, switch on LSAs/paid search and use shared leads only as a stopgap, with five-minute callbacks.
• Goal: backlog reaches 1.5-2 weeks by end of week 3. Crew is producing, not sitting.
• Measure conversion at every funnel stage. Find your real numbers.

Weeks 4-8 (stabilize and tighten)

• Hold backlog in the healthy 2-3 week band. When it dips under a week, spike canvassing hours and mail volume that same week.
• Cluster newly signed jobs geographically; enforce the job-readiness gates so nothing hits the board un-ready.
• Keep a live backfill/rain list so weather never empties the crew.
• Start the review flywheel, ask on every completed job. Pay out the first referrals fast and loud.
• Begin weaning off the most expensive paid leads as warm/organic sources come online.

Weeks 9-12 (compound and decide on crew #2)

• Your reviews, signs, referrals, and repeat-neighborhood presence (the slow clock) should now be generating leads at a fraction of week-1 cost.
• Re-canvass the tracts you sold into; clustered jobs and yard signs make these knocks warm.
• Watch the two triggers: backlog steady at 3-plus weeks AND a real surplus of leads you can't service. If both are true, you're ready to staff crew #2. If not, hold and keep the one crew full and profitable.

What full looks like

For a single residential reroof crew, "full and healthy" is roughly: 3.5-4 installs/week, a 2-3 week backlog, jobs clustered so windshield time stays low, a ready backfill list so rain never idles them, and a lead engine producing your ~25-30 weekly qualified opportunities at a blended cost that's dropping month over month as the slow clock kicks in. Hit that and the crew is no longer a cash risk, it's a profit center, and you've got the playbook to clone it.

The few things new owners get wrong (so you don't)

• Treating "busy" as the goal instead of a backlog number. Measure the gauge. Two-to-three weeks. Defend it.
• Funding only the slow clock. SEO and reviews are real, but they won't pay this Friday's payroll. Fund the fast clock now, plant the slow clock for later.
• Knocking random streets. The yield difference between targeted and blind canvassing is enormous. Knock where roofs are due, age plus storm, not where you happen to be.
• Booking with zero slack. No buffer day, no backfill list, then one rainy Monday cascades into refunds.
• Letting green reps freelance the claims pitch. "I'll get you a free roof" is the fastest way to a complaint, a fine, or worse. Document, estimate, hand off. The homeowner files; the insurer decides.
• Putting un-ready jobs on the board. A signed contract isn't a schedulable job until materials are confirmed and permits are clear.
• Adding crew #2 on hope. Wait for both triggers, sustained backlog and real lead surplus, before you double the payroll.

Filling a new crew's schedule from scratch is not luck and it's not charisma. It's a number you can calculate, channels you deploy in the right order, lists you target instead of spray, and a schedule you protect from weather and your own over-promising. Run the math, point the trucks at the roofs that are actually due, keep the board honest, and the empty Monday yard becomes a problem you solved once and never again.

If the targeting piece is where you keep losing days, that's exactly the gap RoofPredict is built to close: it scores addresses by roof-age range and models storm physics per roof so your canvassers and mailers hit the streets most likely to be due, and it enriches your own list or CRM with those signals so you stop spraying and start routing. Honest about the limits, age is a range, storm is odds, your foreman's eyes still make the call, but it turns the guesswork of "where do we knock Monday" into a ranked route, which is the difference between a crew that's idle and a crew that's full. Build the engine once, and the next crew you stand up fills twice as fast.

FAQ

How many leads do I need to keep one roofing crew busy?

At typical retail conversion rates (about 60% lead-to-appointment, 80% appointment-to-estimate, 35% estimate-to-close), you need roughly 25 to 30 genuine qualified leads per week to keep one residential reroof crew full at about 3.5 to 4 installs per week. If you close higher, for example storm or insurance-driven work at 50 percent plus, the requirement drops to roughly 16 to 18 leads a week. Track your own conversion at each stage and rebuild the math with your real numbers.

What is the fastest way to fill a brand-new crew's schedule?

Targeted door knocking, backed by your existing warm network. Canvassing converts attention into appointments with no ad spend and a same-day feedback loop, and your past customers, suppliers, and property-manager contacts are the highest-closing source you already have. The key is targeting: knock streets where roofs are genuinely due (same build era aging out, or a recent storm footprint), not random streets. Paid search and shared leads can bridge a gap but cost more per signed job.

How big a backlog should a healthy roofing crew have?

Aim for a 2 to 3 week backlog per crew. That is enough that a couple of rain days don't empty the board, but not so much that customers who signed start canceling because you can't start for a month. Treat the backlog like a fuel gauge: when it drops under one week, spike your lead-gen and canvassing; when it climbs past three to four weeks, slow marketing or get ready to add a second crew.

How do I keep a crew from sitting idle on rain days?

Carry a live backfill list of jobs that are ready to install but flexible on date, ideally weather-tolerant work like covered porches, detached garages, repairs, and partials. When a Tuesday rains out, you slide a backfill job up instead of losing the day. Also leave a floating buffer day in each week rather than booking five installs into five days, so a single rainout doesn't cascade into a three-week delay.

When should I add a second roofing crew?

Wait for two signals at once: your backlog sits consistently at three-plus weeks AND you have a real surplus of leads you're already paying to generate but can't service. Adding a crew before the pipeline can feed it just doubles your idle payroll. Adding it too late means refunded deposits and damaged reviews from missed start dates. Let those two numbers, not optimism, set the timing.

Is door knocking worth it compared to buying leads?

For a new crew, usually yes. Canvassing costs labor instead of cash, converts faster, and you keep the relationship rather than racing four other contractors for a shared lead. Bought leads from aggregators are typically sold to several companies at once, so close rates run low and the effective cost per signed job can be high. Use paid and shared leads as a stopgap with five-minute callbacks; build your foundation on targeted canvassing, your database, and referrals.

How do I target the right streets for canvassing and mail?

Look for two signals that stack: roof age (subdivisions built in the same era with original shingles aging out together) and storm exposure (neighborhoods inside a real hail or high-wind footprint). When both line up, your knock-to-appointment rate can double or triple versus blind knocking. You can build these lists from county build-year records and storm reports plus windshield time, or use a tool that scores addresses by roof-age range and models storm physics per roof, like RoofPredict, to rank routes before your crew leaves the yard.

What can my sales reps legally say about insurance at the door?

They can offer to inspect, photograph, and thoroughly document the roof, and write an accurate repair estimate they hand to the homeowner. They must not promise the insurer will approve a claim, promise a free roof, promise the deductible will be waived or absorbed, or offer to handle, negotiate, or fight the claim with the carrier on the homeowner's behalf, that last set is unlicensed public adjusting in most states. The compliant frame: you document and estimate, the homeowner files, and the insurer decides coverage.

What does RoofPredict actually do for crew scheduling?

RoofPredict scores addresses by an estimated roof-age range read from aerial imagery and overlays storm physics modeled per individual roof, then ranks streets and lists so your canvassers and mailers focus on the doors most likely to be due. It can also enrich your own CRM or mailing list with those age and storm signals. Be clear on the limits: roof age is a range, not a build date, and storm modeling gives odds a roof was affected, not proof of damage, so your foreman's inspection still makes the final call. It points the trucks at the right blocks; it isn't a lead-buying service.

How long until a new crew's schedule is reliably full?

With a disciplined ramp, expect a 1.5 to 2 week backlog by the end of week three from canvassing, mail, and your warm network, then stabilizing into the healthy 2 to 3 week band over weeks four to eight. By weeks nine to twelve, the slow-clock channels (reviews, referrals, yard signs, repeat-neighborhood presence) should start lowering your blended cost per job. The fast-clock channels fill the board immediately; the slow clock is what makes keeping it full cheap by month four.