Reduce Roofing Canvasser Turnover by Improving Door Quality

A canvasser quits on a Tuesday afternoon, three streets into a neighborhood your office manager pulled off a county parcel export at 7 a.m. that morning. He knocked forty doors. Eleven people answered. Two of them already had a roofer. One was renting and couldn't make a decision. Seven had roofs that looked fine from the street and felt no reason to talk to him. He set zero appointments, made zero dollars on a draw he now has to pay back, and by the time he texts his crew lead "I'm done," you've already lost the $1,800 you spent recruiting, onboarding, and outfitting him with a polo and an iPad.

The owner's instinct, almost every time, is to blame the rep. He didn't have the grit. He couldn't handle rejection. Door-to-door isn't for everyone. All of that contains a grain of truth and misses the actual mechanism. The rep didn't fail because he was weak. He failed because the doors were bad, and bad doors break good people faster than rejection ever does.

Roofing canvasser turnover sits somewhere between brutal and catastrophic across the industry. It is common for a storm-chasing operation to run through three or four canvassers for every one who lasts a full season, and for the survivors to be the ones who happened to get the good neighborhoods in their first two weeks. The companies that quietly keep their door knockers are not paying dramatically more. They are not running secret motivational programs. They are sending people to better doors, and the retention follows almost mechanically from that one decision. Improve the quality of the door and you reduce the turnover, because the rep gets the early wins that keep him in the truck long enough to actually get good.

This is the relationship most owners never connect. Door quality is a retention lever, arguably the most powerful one you have, and it is almost entirely within your control. The list, the routing, the targeting, the way you stack a rep's first three weeks with winnable conversations rather than impossible ones. Below is how the good operators think about it, the math underneath the people problem, the workflows that fix it, and the places where targeting data earns its keep.

Why Canvassers Actually Quit (It's Not What You Think)

If you exit-interview enough roofing canvassers, a pattern shows up that has very little to do with the reasons they give. People who quit say things like "it wasn't a good fit" or "I found something with better hours" or simply ghost. What actually happened, in case after case, is that the activity-to-reward ratio collapsed and stayed collapsed long enough to feel permanent.

A door knocker can absorb an enormous amount of rejection if the rejection is occasionally interrupted by a win. The human brain is built for variable rewards; it is why slot machines work. A rep who knocks eighty doors and sets two appointments has had a good day, and the two appointments are the payout that makes the seventy-eight no-answers and not-interesteds tolerable. The same rep who knocks eighty doors over two days and sets zero appointments has not had a day twice as bad. He has had a day that is categorically different, because the reward never arrived, and the brain reads a long stretch with no payout as a broken machine. He doesn't quit because he knocked eighty doors. He quits because he knocked eighty doors into a void.

The first-three-weeks cliff

Watch when people leave and you will see the timing cluster. Very few canvassers quit on day two; the novelty and the draw carry them. The dangerous window is roughly days ten through twenty-five. The rep has been out long enough that the newness is gone, the draw is starting to feel like debt rather than income, and he has formed a belief about whether this works. If his first three weeks were spent on a thin, picked-over, or badly targeted list, he has now lived the experience of door-knocking-as-futility and generalized it into a verdict: this doesn't work, or I can't do this. Both are conclusions about reality drawn from a sample you controlled.

The operators who keep people understand that the first three weeks are not a test of the rep. They are a test of the company's ability to put the rep in front of winnable conversations. A new canvasser on a genuinely good list, knocking neighborhoods where a meaningful share of roofs are aging out or storm-worn, will stumble into enough early appointments to build the belief that effort produces reward. That belief is the entire game. Once it is set, the rep will knock through bad days because he knows the good days exist. Without it, no amount of pizza parties or motivational meetings will hold him.

Rejection versus futility

There is a crucial distinction that owners blur: rejection and futility are not the same experience, and only one of them is the job. Rejection is a homeowner saying no, being rude, slamming a door. It stings, but it is a normal, expected, almost respectable part of sales, and reps can be trained to reframe it. Futility is different. Futility is knocking doors where no realistic prospect exists no matter how good the pitch is, because the roofs are new, the houses are rentals, the homeowner already signed with a competitor last week, or nobody is home because the list is built on stale data. Futility is not a sales skills problem. It is a targeting problem wearing a sales skills costume, and you cannot coach your way out of it.

The reason this matters for turnover is that reps misattribute futility as personal failure. A rep who spends a week on roofs that will not need replacing for fifteen years concludes he is bad at sales. He is not. He was given an unwinnable assignment and blamed himself for losing. When you fix door quality, you are doing more than improving conversion numbers; you are removing the source of the false belief that drives your best raw talent out the door before they ever get good.

The Hidden Economics of Turnover

Owners chronically underprice their own turnover because most of the cost is invisible on the P&L. There is no line item called "canvasser we lost." The money leaks out in a dozen small ways that never get summed. Sum them and the number reorganizes your priorities.

What one lost canvasser actually costs

Walk through it concretely. To replace a canvasser you re-run recruiting: job board spend, the hours your sales manager spends screening and interviewing, the no-shows to interviews, the people who accept and never show up to day one. Industry recruiting-cost studies from sources like the Society for Human Resource Management consistently put the cost-per-hire for entry-level roles in the range of a few thousand dollars once you load in advertising and staff time, and door-knocking roles often carry higher costs because the funnel is leakier.

Then you onboard: shadow days where a producing rep slows down to train, draw paid out before the rep generates anything, gear, software seats, background checks where required. Then comes the productivity ramp. A canvasser is not at full output on day one; he climbs a curve over weeks. If he quits at day twenty, you paid for the entire ramp investment and harvested almost none of the plateau production it was supposed to fund.

A rough but honest model: assume $1,500 to $2,500 in hard recruiting and onboarding cost, plus a draw that ran $1,000 to $3,000 ahead of self-funding production, plus the opportunity cost of the territory that sat unproductive while a doomed rep worked it and a replacement got hired. It is entirely reasonable for a single early-departure canvasser to cost a roofing company $4,000 to $8,000 all-in. Run through six of them in a season chasing the three that stick and you have spent $24,000 to $48,000 learning a lesson that better doors would have taught you for free.

The territory-burn cost nobody counts

Here is the cost that is genuinely invisible and genuinely large: a bad-fit canvasser does more than fail to produce, he degrades the territory on his way out. Every door he knocks badly, every neighborhood he works while demoralized and going through the motions, is a door that is now harder for the next rep. Homeowners remember being knocked. A street that has been hit by a checked-out canvasser reading a flat script is a colder street for the next six months than a street that was never knocked at all. You are not only losing the rep. You are spending down the freshness of your best neighborhoods to subsidize a hire that targeting could have saved.

This is why the math is not linear. Turnover does not merely cost you the replacement; it costs youthe burned inventory of good doors, which is the actual scarce resource in a canvassing business. Treat your good neighborhoods like the finite asset they are and the calculus around door quality changes completely.

A simple worked example

Take two crews of four canvassers each, run for one 90-day storm season.

The appointment gap is the obvious story. The quieter story is that Crew A burned seven good neighborhoods staffing a revolving door of demoralized reps, while Crew B preserved most of its inventory and kept the people who were starting to get good. A year from now, Crew B's company has a trained, tenured canvassing team and fresh territory. Crew A's company is starting over, again, and has trained its remaining market to recognize and avoid its brand.

What "Door Quality" Actually Means

Door quality is a vague phrase until you decompose it. A high-quality door is not simply a house with a damaged roof. It is the intersection of several conditions, each of which independently raises the odds that a knock turns into a conversation and a conversation turns into an appointment. Get specific about the components and you can actually build for them.

The five components of a good door

Roof need. The roof is genuinely approaching or past the point where replacement is reasonable, either because it is aging out of its service life or because a storm did measurable work on it. This is the single biggest driver and the one targeting data most directly improves. A roof that is eight years into a thirty-year architectural shingle is a bad door no matter how good your rep is; a roof that is twenty-two years into a twenty-year three-tab, or a roof that sat under softball-size hail three weeks ago, is a fundamentally different conversation.

Owner-occupancy and decision authority. The person who answers can actually say yes. Rentals, absentee-owned homes, and properties in the middle of a sale are structurally worse doors because the human at the door cannot authorize the work. A neighborhood that looks great on roof age but is forty percent rentals will quietly destroy a rep's numbers.

Reachability. Someone is actually home, and the address is real and current. A list built on stale data sends reps to demolished houses, perpetual no-answers, and addresses that have changed hands twice since the export. Reachability is unglamorous and enormously important; the best-targeted list in the world is worthless if half the doors never open.

Competitive freshness. A roofer hasn't already saturated the street. After a storm, the first crews in get the easy conversations; by week four, every homeowner on the block has talked to five companies and signed with one or built a wall of sales resistance. Freshness decays fast and is part of what makes a door good or bad on a given day.

Approachability. Practical, on-the-ground factors: walkable density so a rep can hit many doors per hour, reasonable income mix so financing or self-pay is realistic, safety, and HOA or no-soliciting friction. A technically perfect target neighborhood with houses a quarter mile apart on private drives will exhaust a rep before he gets twenty knocks in.

A good door scores well on most of these at once. The reason most lists produce bad doors is that they optimize for none of them; a raw parcel export is sorted by nothing that matters and contains every rental, every new build, every demolished structure, and every roof with fifteen years of life left, all mixed together with the genuine prospects.

The two roads to a roof being "due"

There are two independent reasons a roof becomes a real prospect, and a strong canvassing program targets both rather than fixating on one.

The first is age. Asphalt shingle roofs, which dominate the residential market, have finite service lives. Three-tab shingles realistically last in the high teens of years; architectural laminates run longer but still age out. The NRCA and shingle manufacturers publish service-life expectations, and the practical reality is that a large share of any mature neighborhood is sitting somewhere on that curve. A roof that is plausibly past or near the end of its service life is a roof whose owner is increasingly likely to be thinking about replacement, dealing with nuisance leaks, or about to be told by an insurer or a home-sale inspector that it needs attention.

The second is storm. Hail and high wind do real, sometimes sudden damage that can move a roof from "fine for years" to "needs replacement now." Hail in particular fractures the mat and knocks the protective granules off asphalt shingles, and high winds lift and crease tabs; the IBHS and NOAA's Storm Prediction Center document where and how severe these events are. A roof that took a significant hail or wind event recently is a high-need door regardless of its age, and the homeowner often does not yet know the extent of the damage because hail bruising is not always visible from the ground.

The operational point is that age and storm are different signals that identify different doors, and the best targeting layers them. Age tells you which roofs are wearing out on their own schedule; storm tells you which roofs just got pushed off schedule. A neighborhood that is both mature in age and recently storm-struck is the highest-quality door inventory you will find, and it is exactly the kind of overlap a smart list-build is designed to surface.

Building a Targeting System That Produces Good Doors

Knowing what a good door is gets you nowhere until you can systematically generate lists of them. Most roofing companies build canvassing lists in one of three ways, and the first two are why their canvassers quit.

The three ways companies build lists

The spray method. Pick a zip code, send reps to knock everything. This is the default for new and undisciplined operations. It guarantees that the majority of doors a rep knocks are bad doors, because in any given area most roofs at any given moment are not due. The spray method is the single largest manufacturer of canvasser turnover in the industry. It feels like activity and produces futility.

The storm-only method. Wait for a storm, pull a hail or wind swath map, send everyone into the swath. This is much better than spraying because the storm pre-qualifies for need, and it is why storm chasers can run brutal turnover and still survive; the doors are good enough during the brief fresh window that even a leaky people-system produces. The weakness is that it is feast-or-famine, intensely competitive in the first weeks, and it leaves your team idle and your reps starving between events.

The layered-data method. Build lists from the intersection of multiple signals: roof age estimates, storm exposure, owner-occupancy, and reachability. This is what the operators with low turnover do. It produces good doors year-round, rather than only in the two weeks after a storm, and itlets you keep a tenured team busy and producing through the slow stretches that otherwise bleed your roster dry.

The layered method is the goal. The rest of this section is how to build toward it.

Layering the signals

Think of list-building as a series of filters applied in sequence, each one removing bad doors before a rep ever sees them. The order matters because each filter is cheaper or more expensive to apply.

Start with the cheap, high-leverage filters. Remove anything that is structurally a bad door regardless of roof condition: known rentals and absentee-owned parcels where you can identify them, brand-new construction, commercial where you only do residential, and addresses that public records flag as recently sold or in foreclosure limbo. This step alone, done off basic parcel and ownership data, can strip a meaningful fraction of dead doors out of a raw export.

Next, layer roof need. This is where you separate the mature, due-soon roofs from the ones with a decade of life left, and where you flag recent storm exposure. Historically this was the hard part; a sales manager would eyeball satellite imagery house by house, or simply guess based on neighborhood build-year, which is crude because roofs get replaced at different times and a 1998 subdivision today contains original roofs, ten-year-old reroofs, and brand-new ones side by side. Per-roof signals, rather than per-neighborhood guesses, are what make this filter actually work.

Then layer reachability and freshness. Suppress addresses you have already worked recently, addresses other crews are working, and known bad-data addresses. Append phone or mailing data if your strategy includes multi-touch, so canvassing is not the only contact.

What comes out the other end is a list where most doors clear most of the quality bar. That is the entire objective: not a perfect list, but a list where a rep's default door is a decent door rather than a long shot.

Where RoofPredict fits

The filter that is hardest to do by hand at scale is roof need, and it is the one that moves door quality the most. This is the specific problem RoofPredict is built for. It analyzes aerial imagery to estimate a roof-age range for individual addresses, house by house, and it models storm physics per roof so you can see which roofs a given hail or wind event likely worked over rather than just which roofs fell inside a coarse swath polygon. The output is a per-address signal you can use to rank and filter, so your canvassers spend their hours on the roofs that are aging out and the roofs the storm wore down, rather than on a parcel export sorted by nothing.

It is worth being precise about what this is and is not, because over-promising targeting data is its own way to burn out a team. RoofPredict gives you a roof-age range, not a roof's exact install date; aerial imagery cannot read a birth certificate off shingles, and the honest output is a band like "roughly eighteen to twenty-four years" that tells you a roof is in the danger zone without pretending to a precision it does not have. The storm modeling produces odds and exposure estimates, not proof that a specific roof is damaged; it tells you which roofs were most likely affected, and a rep or inspector still has to get on the roof to document what is actually there. Used correctly, it does not replace a canvasser's judgment or an inspector's eyes. It replaces the part of the job that was previously a guess, which is which doors are worth knocking in the first place.

The practical effect on turnover is straightforward. When the roof-need filter is doing real work, a higher fraction of the doors your rep knocks are genuine prospects, the activity-to-reward ratio improves, the early wins show up, and the rep stays past the day-twenty cliff. The data does not motivate anyone. It just stops sending good people to roofs that will not need replacing for a decade. It also enriches lists and CRM records you already own, so the contractor's existing mailing list or canvassing territory gets a roof-age and storm layer added to it rather than being thrown away. RoofPredict is not a lead-buying service and does not hand you customers; it ranks the doors, routes, and lists you are already going to work so your team works the right ones first.

Routing and Density: The Other Half of Door Quality

A good list still produces a bad day if the route is bad. Two reps can be handed the same high-quality target addresses and have completely different experiences depending on how those addresses are sequenced into a day of walking. Routing is the part of door quality that lives in the field, and it is consistently underrated.

Knocks per hour is the number that matters

The metric that determines whether a canvasser has a productive day is knocks per hour at decent doors. A rep who can knock thirty good doors an hour because they are dense and walkable will out-produce a rep knocking twelve good doors an hour spread across a sprawling subdivision, even though the second rep's individual doors might be marginally better. Density is a multiplier on door quality. The best target neighborhood in your market is a worse assignment than a merely good one if the houses are so spread out that a rep spends his day driving between them.

This is why routing should optimize for tight clusters of good doors, not for the single best doors scattered across a county. Group your filtered target addresses geographically, hand a rep a compact walkable block of them, and let him stay on foot. A canvasser who is walking and knocking is producing; a canvasser who is driving or circling for parking is bleeding the day away. The companies with the best per-rep numbers are almost always the ones who solved density, not the ones who found magically better individual doors.

Time-of-day and day-of-week realities

Door quality has a time dimension that routing has to respect. The same door is a good door at 6 p.m. on a weekday and a bad door at 10 a.m. when the owner-occupant is at work. Reachability collapses during work hours in owner-occupied neighborhoods, which means a rep sent to a great residential list at 10 a.m. on a Tuesday will experience it as a bad list, because nobody is home. Saturdays and weekday early evenings are when owner-occupied doors actually open.

Smart operators route accordingly: lower-reachability windows go to areas with more retirees or work-from-home density if they have the data to know where those are, while the prime evening and weekend windows are reserved for the highest-need owner-occupied targets. A rep who only ever gets the prime windows on good lists has a very different career than one who keeps getting sent to ghost-town doors at the wrong time and concludes the job doesn't work.

A sample daily route build

Here is a concrete way a crew lead can assemble a canvasser's day from a layered target list:

1. Pull the filtered, high-need addresses for the assigned area, already stripped of rentals, new builds, recently worked, and low-reachability records.
2. Cluster them into walkable groups of roughly 40 to 70 doors that a rep can cover on foot in a two-to-three hour push.
3. Assign the densest, highest-need cluster to the prime evening window, and a secondary cluster as overflow.
4. Give the rep a clear physical or digital boundary so he is not improvising routes and wandering into bad doors out of indecision.
5. Log every knock and outcome so the next day's route can suppress no-answers for a callback pass and permanently remove dead doors.
6. Reserve the highest-density, freshest cluster for any brand-new rep, deliberately, to engineer early wins.

That last step is the retention move hiding inside a routing process. You do not give your new rep the leftover doors nobody wanted. You give him the best winnable cluster you have, on purpose, because his first three weeks decide whether he becomes a tenured producer or another line in your turnover spreadsheet.

Stacking the First 30 Days for Retention

If door quality is the lever, the new-canvasser onboarding window is where you pull it hardest. The single highest-return retention intervention available to a roofing company costs nothing but discipline: deliberately front-load a new rep's first month with the best doors you have.

Engineer the early win

A new canvasser needs a win in his first day or two, not his third week. The win does not have to be a closed deal; for a canvasser it is a set appointment, a genuinely interested homeowner, a roof where his inspection found real damage. That win is the proof, delivered to the rep's own nervous system, that the activity produces reward. It builds the belief that carries him through the inevitable bad days.

Most companies sabotage this by giving new reps the worst doors. The logic, usually unspoken, is that veterans have earned the good neighborhoods and rookies should pay dues on the picked-over ones. This is exactly backwards from a retention standpoint. Veterans have already formed the belief that effort produces reward and can withstand a stretch of bad doors; rookies have not and cannot. You are spending your best doors on the people who least need them and starving the people whose retention depends on them. Flip it. Veterans can grind marginal doors because they know the good days are real. Rookies need the good doors precisely because they don't know that yet.

A 30-day door-quality ramp

Structure the first month as a deliberate quality curve:

The days 11 through 20 row is the one to watch obsessively, because it is where most quitting happens. A rep whose first ten days went well can still crater in week two if the doors suddenly turn marginal at the exact moment his confidence is most fragile. Keep the doors decent through that window even if it means a veteran eats a thin patch.

Coach attribution, not only technique

The other half of the first 30 days is teaching reps to attribute outcomes correctly, because the belief that drives turnover is a story reps tell themselves about why a bad day happened. A rep who has a bad day and thinks "I'm not cut out for this" is on his way out. A rep who has the same bad day and thinks "that block was mostly rentals and it was 10 a.m., this is a list-and-timing problem, the next block will be different" is on his way to tenure. The difference is attribution, and you can coach it.

This is easier to coach honestly when the lists are actually good, because then the attribution is also true. You cannot tell a rep "it's the doors, not you" with a straight face if you are still spraying zip codes; he will correctly sense that you are managing his morale rather than fixing his problem. Good targeting earns you the right to coach attribution, because the doors really are good and the bad days really are noise rather than signal. This is the deep reason door quality and retention are linked: quality doors make the truthful, retention-building story the actual true story.

Compensation Plans That Don't Punish Good Reps

Door quality and compensation interact in ways that quietly determine who stays. A comp plan can either work with good doors to retain people or fight against them and drive people out even when the doors are fine. Most roofing canvasser comp plans were designed by someone who never knocked, and it shows.

The draw trap

The most common structure is a draw against commission: the rep gets a weekly draw, say $600, which is an advance he pays back out of future commissions. On good doors this works; the rep out-earns the draw quickly and the draw is just smoothing. On bad doors it becomes a debt trap. The rep falls behind the draw, watches his "balance" grow, and experiences his own job as a source of mounting debt. There are few faster ways to make someone quit than to have them owe you money for the privilege of being demoralized.

The draw trap and the door-quality problem compound each other viciously. Bad doors mean low production; low production against a draw means growing debt; growing debt means the rep is now financially punished for the company's targeting failure. He didn't choose the bad list, but he is the one whose paycheck absorbs its consequences. When reps in this situation quit, owners often say "he couldn't hack the commission grind," when what actually happened is the company built bad lists and then made the rep pay for them.

Structures that align with retention

A few structural choices make comp plans far more retentive, especially in the critical early window.

Non-recoverable draw during onboarding. For the first few weeks, make the draw a guarantee, not an advance to be clawed back. This removes the debt-spiral risk during exactly the window when the rep is most likely to quit and least able to control his outcomes. It costs you some money on reps who wash out, and it saves you the much larger cost of losing reps who would have made it if they hadn't been buried in draw debt at day fifteen.

Pay for activity quality, not only outcomes, early on. A new rep cannot control whether a homeowner says yes, but he can control whether he knocked the assigned doors, logged outcomes honestly, and ran the inspection process correctly. Paying a small bonus for clean, complete activity during onboarding rewards the behaviors that lead to production before production itself is reliable, and it keeps reps engaged through the ramp.

Tier commissions so the early appointments pay. If your plan only pays meaningfully on closed-and-built jobs, a canvasser who sets appointments but hands them to a closer waits weeks for a payout and may quit before it arrives. Pay something real and fast on the set appointment or the qualified inspection, so the reward lands close in time to the activity that earned it. Variable rewards work when they are frequent and proximate, not when they are distant and rare.

Never penalize a rep for good doors going to a teammate. If two reps are competing for the same neighborhoods, your routing is creating internal conflict that the comp plan then amplifies. Assign clean, non-overlapping territory so that a rep's earnings depend on his work, not on whether he got to a block before his coworker.

A worked comp comparison

Consider a new rep over his first three weeks under two plans, on a decent-but-not-spectacular list where he sets nine appointments that eventually become four sold jobs.

Under a pure recoverable draw of $600/week that only pays on built jobs, by end of week three he has drawn $1,800, has zero commission paid because nothing has been built yet, and sees an $1,800 debt against his name. His subjective experience: "I've worked three weeks and I'm $1,800 in the hole." High quit risk regardless of how the jobs eventually turn out.

Under a non-recoverable $600 onboarding guarantee plus a $75 per-qualified-inspection spiff, he has earned $1,800 guaranteed plus, say, $525 in inspection spiffs, with the built-job commissions still coming later as upside. His subjective experience: "I've made about $2,300 in three weeks and I have four jobs in the pipeline." Low quit risk, and the eventual commissions are gravy that deepen his commitment.

Same rep, same doors, same nine appointments. One plan retains him and one plan loses him. The doors set the ceiling on what is possible; the comp plan determines whether the rep survives long enough to reach it.

The Inspection-and-Documentation Workflow (Done Right)

Much canvassing in roofing exists to set up a roof inspection, and a great deal of turnover and legal trouble both flow from how that inspection step is run. Reps need a clean, repeatable, compliant workflow, both because it converts better and because it keeps them and the company out of trouble that can end a career or a business.

What the canvasser is actually doing at the door

The honest framing of the canvasser's job is to identify roofs that may be due based on age or recent storms, offer the homeowner a thorough inspection and documentation of their roof's condition, and, where damage exists, prepare an accurate repair estimate the homeowner can use. That is a clean, defensible, valuable service. It is also the framing that keeps everyone on the right side of the line.

Where companies get into trouble is when canvassers, hungry for a yes and often coached badly, start saying things they are not allowed to say. This is worth teaching explicitly to every rep, because a single canvasser making prohibited claims on a doorstep can expose the company to regulatory action under unfair-trade-practice and unlicensed-public-adjusting rules that vary by state and are enforced by state departments of insurance and attorneys general.

The do-not-say list, taught plainly

Train every canvasser that they may document the roof thoroughly, photograph and describe what they actually observe, and prepare an accurate, Xactimate-aligned repair estimate for the scope of work. The homeowner then files any insurance claim themselves, and the insurer decides what is covered. The canvasser states facts about the roof and the proposed repair scope; the canvasser does not touch the claim.

What a canvasser must never do, for a fee or as part of the sales pitch:

• Negotiate, adjust, or "handle" the homeowner's insurance claim. That is the homeowner's and the insurer's role; doing it for a fee is unlicensed public adjusting in most states.
• Interpret the homeowner's policy or tell them what is or isn't covered. They are not the insurer and not a licensed adjuster.
• Promise a specific payout, a guaranteed approval, or that "insurance will definitely cover this." Coverage is the insurer's decision and cannot be promised.
• Promise the deductible will be waived, absorbed, eaten, or made to disappear. Offering to cover a homeowner's deductible is illegal in many states and is a fast route to fraud exposure.
• Advertise or promise a "free roof." The roof is not free; it is paid for by a claim the homeowner files and the insurer decides, with the deductible owed by the homeowner.
• Represent the homeowner against the insurer in any way. The canvasser works for the roofing company and prepares documentation; he does not act as the homeowner's advocate in the claim.

Teaching this plainly is more than compliance hygiene; it is a retention and reputation asset. Reps who are given clear, honest, legal language to use are more confident at the door than reps who are winging it with half-remembered claims they suspect are too good to be true. And a company that does not chase "free roof" promises builds a market reputation that makes every future door warmer.

A clean five-step field workflow

1. Identify and approach. The rep approaches a targeted address with a specific, honest reason: the neighborhood's roofs are at an age where issues are common, or a recent storm passed through, and he is offering free inspections to document condition.
2. Offer the inspection. No promises about outcomes. The offer is to get on the roof, document its actual condition with photos, and tell the homeowner the truth about what is there.
3. Document thoroughly. Photograph damage and wear with consistent, dated, address-tagged images. Note hail bruising, granule loss, wind creasing, flashing condition, age indicators. Document what is, not what the homeowner hopes is there.
4. Prepare an accurate estimate. Where genuine damage or end-of-life condition exists, write an honest, line-item repair estimate aligned to standard pricing the homeowner can hand to their insurer or use for a self-pay decision.
5. Hand it off correctly. Give the homeowner the documentation and estimate. The homeowner decides whether to file a claim or proceed self-pay; the insurer decides coverage. The company stays in its lane as the documenter and contractor.

A rep who runs this workflow cleanly converts well, sleeps well, and does not expose the company. A rep who was never taught it improvises, over-promises, and becomes both a turnover risk and a liability.

Measuring Door Quality So You Can Manage It

You cannot improve door quality if you are not measuring it, and most companies measure reps without ever measuring doors. The fix is to track metrics that separate the contribution of the list from the contribution of the person, so you stop firing reps for list problems and stop blaming lists for rep problems.

Metrics that isolate door quality from rep skill

The core idea is to look at the same metric across many reps on the same doors, and the same rep across different doors. Patterns emerge that tell you which is which.

Contact rate (doors answered / doors knocked). This is mostly a property of the list and the timing, not the rep. If contact rate is low across all reps on a given area, you have a reachability or timing problem, not a people problem. If one rep has a low contact rate while everyone else on the same area is fine, look at that rep's hours and effort.

Conversation-to-appointment rate (appointments set / doors answered). This is more about the rep's skill and the door's underlying need, blended. Low across all reps on high-need targets suggests a pitch or training problem. Low on a specific area despite good contact suggests the roofs are not actually as due as the list claimed, which is a targeting-data problem.

Inspection-find rate (real damage or end-of-life found / inspections done). This is a clean read on roof-need targeting. If your reps are getting on roofs and consistently finding genuine wear or storm damage, your need-targeting is working. If inspections keep coming back clean, your roof-age and storm signals are off, and your reps are wasting their best asset, the inspection, on roofs that don't need it.

Per-area appointment yield over time. Track how many good appointments an area produces and watch for decay, which tells you freshness is spent and it's time to rotate. This protects your inventory of good neighborhoods, the scarce asset turnover quietly destroys.

A door-quality scorecard

Maintain a simple scorecard at the area level, updated as reps work it, so routing decisions are made on data rather than vibes.

This scorecard turns door quality from a feeling into a managed input. Maple Ridge goes to your rookies because it is prime and fresh. Cedar Glen's low inspection-find rate is a signal that the roof-age data was wrong, not that the reps are bad, so you fix the list rather than the people. Oak Park's roofs are good but nobody's home during the day, so you re-time it instead of abandoning it. Birchwood is burned and needs to rest. None of those decisions are available to a company that only tracks individual rep close rates.

Closing the loop with targeting data

The inspection-find rate is the metric that should feed directly back into your targeting. When reps get on roofs and report what they actually find, you are generating ground truth about how well your roof-age and storm signals predicted real need. Feeding that back, whether you do it manually or through a data partner, is how the whole system gets sharper over time. A roof-age range that consistently corresponds to real end-of-life conditions when reps inspect is a range you can trust to send your people toward; a signal that keeps producing clean inspections is one to fix before it burns more reps. The companies that compound their advantage are the ones that treat every inspection as a data point about door quality, not only as a shot at one sale.

Putting It Together: A Retention System, Not a Pep Talk

Step back and the pieces form a single system, and the system is the point. Canvasser turnover is not solved by motivation, harder hiring, or thicker skins. It is solved by raising the quality of the doors your people knock, and then building the comp plan, the onboarding, the routing, and the measurement to compound that advantage.

The causal chain is worth stating one more time because it reorganizes where you spend effort. Better doors produce a better activity-to-reward ratio. A better ratio produces early wins. Early wins build the belief that effort pays. That belief carries reps through the day-twenty cliff. Reps who clear the cliff get good. Reps who get good stay, produce, and stop costing you $5,000 each to replace while burning your best neighborhoods on the way out. Every link in that chain traces back to door quality.

Here is the order of operations for a company that wants to actually fix this:

1. Stop spraying. Kill the zip-code-and-pray lists immediately; they are the largest single source of your turnover.
2. Layer your targeting. Filter out structurally bad doors, then add real per-roof need data so age and storm signals decide which doors your reps knock. This is where roof-age range and per-roof storm modeling from a source like RoofPredict does the heavy lifting that hand-eyeballing imagery never could at scale.
3. Route for density and timing. Cluster good doors into walkable pushes and match areas to the right time windows so reachability holds.
4. Front-load rookie doors. Give new reps your best fresh clusters on purpose, and protect the doors through the days-11-to-20 cliff.
5. Fix the comp trap. Use non-recoverable onboarding guarantees and fast, frequent rewards so reps aren't punished with draw debt for the company's list quality.
6. Run a clean, compliant inspection workflow. Document and estimate honestly, teach the do-not-say list, and keep everyone in their legal lane.
7. Measure doors, not only people. Track contact rate, inspection-find rate, and per-area yield so you fix lists when lists are the problem and coach reps when reps are.

None of this requires you to be a better motivator than you are today. It requires you to treat door quality as the operational input it is, and to stop sending good people to roofs that were never going to need replacing.

If the place you want to start is the highest-leverage one, start with the doors themselves. Knowing which roofs in your market are aging out and which ones a storm actually wore down, house by house, is the difference between a list that retains your team and a list that grinds them up. RoofPredict exists to give you exactly that signal: a roof-age range and a per-roof storm read for the addresses you are already going to work, so your canvassers spend their hours on the roofs most likely to be due. It will not knock the door or close the deal, and it deals in ranges and odds rather than certainties, but it answers the one question that decides whether your canvassers stay: is this a door worth knocking? Get that right and the turnover problem you have been treating as a people problem starts to dissolve into what it always was, a door-quality problem you can fix.

FAQ

Is high roofing canvasser turnover really a door-quality problem and not a hiring problem?

For most companies, yes. Hiring matters, but the dominant driver of early turnover is the activity-to-reward ratio reps experience in their first three weeks. Reps sent to bad doors, meaning roofs with years of life left, rentals, no-answers, and picked-over streets, experience futility and conclude they can't do the job, when the real problem was the assignment. Improve the doors and the same reps who would have quit tend to stay, because they finally get the early wins that build the belief that effort produces reward.

What actually makes a door high-quality for roofing canvassing?

Five things at once: genuine roof need (the roof is aging out or was recently storm-worn), owner-occupancy so the person at the door can say yes, reachability so someone is actually home and the address is current, competitive freshness so the street isn't already saturated, and approachability meaning walkable density and a realistic income mix. A raw parcel export optimizes for none of these, which is why spray lists produce so many bad doors.

When in their tenure do canvassers usually quit, and why does timing matter?

Quitting clusters around days 10 to 25, not day two. The novelty and the draw carry reps early; the danger window is when newness fades, the draw starts feeling like debt, and the rep forms a verdict about whether the job works. If those weeks were spent on bad doors, the verdict is negative and based on a sample you controlled. Deliberately keeping doors good through that window is the single highest-return retention move available.

How much does losing a canvasser actually cost?

More than the visible numbers. Hard recruiting and onboarding costs often run $1,500 to $2,500, plus a draw that ran ahead of production, plus the opportunity cost of an unproductive territory. All-in, a single early departure commonly costs $4,000 to $8,000. The hidden cost is territory burn: a demoralized rep degrades the neighborhoods he works on his way out, spending down your scarce inventory of fresh, good doors and making them colder for the next rep.

Should I give my best doors to new reps or to my veterans?

To new reps, deliberately. The common instinct to make rookies pay dues on picked-over doors is backwards for retention. Veterans have already formed the belief that effort pays and can grind marginal doors; rookies have not and will quit without early wins. Front-load a new rep's first 30 days with your best fresh, dense, high-need clusters, and protect the quality through the days-11-to-20 cliff even if a veteran has to eat a thin patch.

How does roof-age and storm data reduce turnover specifically?

It improves the activity-to-reward ratio at its source. By estimating a roof-age range per address from aerial imagery and modeling storm exposure per roof, targeting data lets you filter out roofs with a decade of life left before a rep ever knocks them. More of the doors a rep knocks are genuine prospects, early appointments show up sooner, and reps clear the cliff. The data doesn't motivate anyone; it just stops sending good people to roofs that won't need replacing for years. Note that roof age comes as a range, not an exact install date, and storm modeling gives odds and exposure, not proof, so reps still inspect to confirm.

What's the difference between a draw and a non-recoverable guarantee, and which retains better?

A recoverable draw is an advance the rep pays back from future commissions; if doors are bad and production lags, the rep accrues debt and often quits feeling buried. A non-recoverable guarantee during onboarding is money the rep keeps regardless. Guarantees retain far better in the early window because they remove the debt-spiral risk during exactly the period reps are most likely to quit and least able to control outcomes. Pair it with fast, frequent rewards like a per-qualified-inspection spiff so payouts land close in time to the work.

What can a canvasser legally say about insurance at the door?

A canvasser can offer to inspect the roof, document its actual condition with photos, and prepare an accurate repair estimate the homeowner can use. The homeowner files any claim and the insurer decides coverage. A canvasser must not negotiate or handle the claim, interpret what the policy covers, promise a specific payout or guaranteed approval, promise to waive or absorb the deductible, advertise a free roof, or represent the homeowner against the insurer. Those actions can constitute unlicensed public adjusting or unfair trade practices under state law. Teaching this do-not-say list plainly protects the company and actually makes reps more confident at the door.

How do I tell whether a bad day was the rep's fault or the list's fault?

Measure doors, not only people. Track contact rate (mostly a list-and-timing signal), inspection-find rate (a clean read on roof-need targeting), and per-area appointment yield. If a metric is poor across all reps on the same area, it's a list, timing, or targeting problem; if it's poor for one rep while others are fine on the same area, it's a rep problem. This separation stops you from firing people for list failures and from blaming lists for genuine coaching needs.

How often should I rotate or rest a neighborhood?

Watch per-area appointment yield over time and rotate when it decays. Competitive freshness is a perishable asset; after a street has been heavily knocked, especially in the weeks following a storm, homeowners build resistance and the doors get materially colder. A neighborhood that has been worked out should rest, often around six months, before you send reps back. Tracking freshness on an area-level scorecard lets you protect your inventory of good doors rather than burning it to subsidize turnover.