How to Divide Roofing Sales Territory Between Reps Fairly

Two of your best reps stop talking to each other. One thinks the other got handed the good side of town. A green hire quits in week six because every door on his map already has a five-year-old roof and a 'No Soliciting' sticker. Meanwhile a hailstorm clips the northeast corner of your service area and three reps suddenly want the same forty streets.

That is what an unfair territory split actually looks like day to day. It rarely shows up as a single blowup. It leaks out as resentment, as reps quietly working outside their lines, as your closer spending Saturdays driving instead of knocking, and as turnover you blame on "they just weren't hungry." Plenty of them were hungry. They were handed a map nobody could win on.

Dividing territory fairly is not about drawing pretty boundaries on Google Maps. It is about making sure every rep has roughly the same shot at the same money for the same effort, that the math behind their map is visible to them, and that you can re-cut the lines when a storm or a hire changes the picture without starting a war. Below is the operating model we would actually run, with the numbers, the worked examples, the edge cases, and the mistakes that quietly wreck good teams.

What "fair" actually means in roofing (and what it doesn't)

Fair does not mean equal square miles. A rep with twenty square miles of subdivisions built in 2019 has a worse territory than a rep with four square miles of 1998 split-levels that took a hailstorm last spring. Geography is the least useful way to slice roofing.

Fair also does not mean equal door counts. Ten thousand doors of brand-new construction is a worse hand than two thousand doors of aging asphalt in a neighborhood that already trusts your yard signs.

What fair means is roughly equal expected opportunity per hour of selling. Two things drive that:

1. The supply of roofs that are genuinely due — roofs old enough or storm-worn enough that a replacement or repair is a real conversation, not a cold pitch to someone who reroofed last year.
2. The cost to work them — drive time, knock density, gate codes, HOA restrictions, and how saturated the area already is by you and by competitors.

If you balance those two things across reps, you have a fair split. If you balance acreage or doors, you have a tidy-looking map and a morale problem.

There is a third thing fairness has to survive: change. A storm, a new hire, a rep who quits, a competitor who moves in. A division that is fair on January 1 and untouchable for twelve months is not fair by April. Build the re-cut rules into the system from day one or every adjustment will feel like a punishment to whoever loses streets.

The fairness test you can say out loud

Before you assign anything, write one sentence you would be willing to read to the whole team:

"Every territory was built to give the same expected jobs per knocking day, measured by roof age, storm exposure, and drive time — and here is the scorecard for each one."

If you cannot say that with a straight face, you have not divided fairly yet. The rest of the work is making that sentence true and showing the scorecard.

The five inputs that should drive every line you draw

Most owners split territory on one variable — usually ZIP codes or "north of the highway / south of the highway" — because it is easy. Easy is why it is unfair. A roof that is due does not care about a ZIP boundary. Here are the five inputs that actually predict where the money is, in rough order of importance for residential reroof and storm work.

1. Roof age supply (the single biggest factor)

An asphalt shingle roof in most of the country is a real replacement conversation somewhere around the 15-to-25-year mark, depending on material, ventilation, slope, and sun exposure. A neighborhood platted in 2015 has almost no due roofs in it today. A neighborhood built in 1999 is in its prime. A 1970s neighborhood may be on its second or third roof, so "year built" alone lies to you — you need the age of the current roof, not the house.

This is the part most territory maps get blind. County records and Zillow show year built, not the date of the last reroof. A house built in 1985 that was redone after a 2019 storm looks ancient on paper and is the worst door on the street. You want a read on the current roof's age as a range, neighborhood by neighborhood, before you draw a single line.

2. Storm exposure, scored per roof — not per map pin

Hail and wind do not fall evenly across your service area, and they absolutely do not fall evenly within a single neighborhood. A hail core can shred the windward, steep-pitched south faces on one block and barely mark the sheltered roofs two streets over. A straight-line wind event lifts shingles on exposed ridgelines and leaves valleys alone.

A basic storm map tells you a county got hail. That is not enough to balance territory, because two reps can both be "in the storm ZIP" and have wildly different odds of real, documentable damage on their actual streets. You want storm exposure modeled closer to the roof level — direction, intensity, the roof's pitch and orientation — so you are balancing worn-out roofs, not map shading.

3. Knock density and drive time

Two territories with the same number of due roofs are not equal if one is tight subdivisions and the other is rural five-acre lots. Knocks per hour is the hidden tax. A rep doing 25 quality conversations a day in a dense neighborhood will out-earn a rep doing 9 in a spread-out one even if the spread-out map has more total due roofs.

Rule of thumb: estimate effective selling hours, not clock hours. A territory with a 25-minute commute each way and lots between is bleeding two-plus selling hours a day before anyone knocks.

4. Saturation and competitive density

A neighborhood you have already worked hard, with eight of your signs up and three jobs in production, is partly spent — and also partly warm. A neighborhood three competitors are currently canvassing is contested. Both change the expected yield. Track which areas you have recently saturated (yours) and which are crowded (theirs), and weight accordingly.

5. Relationship equity (the book)

Past customers, old estimates that never closed, referral clusters, and warranty callbacks are money sitting in a territory before anyone knocks a cold door. If one rep inherits the side of town where you have done 200 jobs over ten years and another inherits raw ground, that is a real and unfair head start unless you account for it. We will handle this directly in the compensation and house-account section, because it is the input that causes the most fights.

Putting the inputs in one place

You do not need a data-science team to use these. You need to stop pretending a ZIP map captures any of them.

A step-by-step workflow to cut territory the first time

Here is the actual sequence. Block out a half-day for the first pass; it gets faster every quarter after that.

Step 1 — Define the whole serviceable area honestly

Draw the outer boundary of where you will actually send a crew without the drive killing the margin. Be honest about the edges. If a job 45 minutes out costs you a half-day of production travel, it may belong to nobody, or to a single "outer ring" rep who knows the deal going in. Map the real working area, not the aspirational one.

Step 2 — Break the area into small "bricks," not big zones

Do not try to draw three or four big territories directly. Break the whole area into small, neutral building blocks — call them bricks. A brick might be a single subdivision, a census-block-group-sized chunk, or a cluster of 150 to 400 homes. Small bricks are the secret to fairness, because you assemble territories out of bricks and can shift one brick at a time later without redrawing the world.

Step 3 — Score every brick

Give each brick a simple opportunity score. You do not need a PhD model. A workable scorecard:

• Due-roof density (0–40): share of homes with a roof in the due age range, scaled by how many homes.
• Storm exposure (0–25): recent, per-roof-modeled hail/wind likelihood. Zero in a calm stretch, high after a real event.
• Workability (0–20): knock density and drive time. Tight and close scores high; rural and far scores low.
• Openness (0–10): inverse of saturation/competition. Fresh ground scores high.
• Book equity (0–5): past customers and open estimates already in the brick.

That is a 100-point brick. The exact weights are yours to tune — the point is to score the same way for every brick so the comparison is honest.

Worked example, three bricks:

Maple Grove and the County Road corridor are close in total value but get there completely differently — one on age, one on a storm. Stoneridge is a trap that looks fine on a door-count map and is nearly worthless today. This is exactly what acreage-based splitting hides.

Step 4 — Assemble balanced territories from bricks

Now stack bricks into territories so the total scores are within ~10% of each other, while keeping each territory geographically contiguous (a rep should not be driving across two other reps' areas to reach an outlier brick).

A clean way to do it with three reps and, say, thirty bricks: sort bricks by score, then snake-draft them into three contiguous piles, balancing totals as you go. You will end up nudging a few bricks by hand to keep each pile connected and within the 10% band. Document the final score of each territory — that is your fairness scorecard.

Step 5 — Sanity-check effective selling hours

Total opportunity can be balanced while one rep still has a brutal commute eating their day. Estimate effective selling hours per territory (clock hours minus travel and dead drive between bricks). If they are lopsided, swap a far brick for a near one between the two reps and re-score. Balance both opportunity and reachable hours.

Step 6 — Write it down and read it to the room

Publish the scorecard. Each rep sees their territory's total, the breakdown, and everyone else's totals. Transparency is the entire game. A rep who can see that their 312-point territory is within 6% of the next rep's 331 will not believe they got robbed. A rep handed a map with no math behind it will always suspect they did.

Three models for splitting, and when each is fair

There is no single correct structure. The fair one depends on your team's tenure mix, your market's storm profile, and how you pay. Here are the three that actually work, with the tradeoffs.

Model A — Fixed exclusive territories

Each rep owns a defined area, exclusively, for a set period (usually a quarter). Best when you have tenured reps who want to farm an area, build referral density, and own their signage footprint.

• Fair when: reps are roughly equal in skill and the bricks are scored and balanced as above.
• Breaks down when: a storm massively over-loads one territory, or skill gaps are wide (a rookie cannot work a prime area as well as a vet, so the "equal" map produces unequal income and the rookie still feels cheated — or the vet is wasted on a weak map).
• Guardrail: quarterly re-cut with published rules, and a storm-override clause (below).

Model B — Rotating zones

Reps rotate through a set of zones on a schedule (e.g., weekly or by campaign). Best for newer teams, for canvassing-heavy outfits, and for keeping any one rep from hoarding the best ground.

• Fair when: zones are unequal and you cannot perfectly balance them — rotation spreads the good and bad ground over time so it evens out.
• Breaks down when: reps cannot build relationship density (you knock a neighborhood once and leave, losing the referral compounding), and follow-up gets messy because the rep who set the appointment is now three zones away.
• Guardrail: keep the follow-up and the job with the rep who originated it even after they rotate out (origination credit, covered below), so rotation does not punish good first contact.

Model C — Tiered / skill-weighted assignment

You intentionally give your strongest closers the highest-opportunity bricks and your green reps lighter, more forgiving ground — and you balance fairness through compensation and ramp, not through equal maps. Best for teams with a wide skill spread.

• Fair when: the comp plan accounts for it — e.g., the rookie's lighter territory comes with a higher commission rate or a ramp guarantee, while the vet's prime territory carries a standard rate and higher volume expectation. Everyone nets a fair shot at a living.
• Breaks down when: you tier the map but pay everyone the same rate — now the rookie is structurally underpaid and will leave.
• Guardrail: make the tiering and the offsetting comp explicit and written. Hidden tiering is the single most resented thing in roofing sales.

Most healthy mid-size roofers run a blend: exclusive territories for tenured reps (Model A), a rotating "new hire" pool of bricks for ramping reps (Model B), and skill-weighting baked into who gets the prime ground (Model C).

Worked example: splitting one market three ways

Let us make it concrete. You run residential reroof and storm work. You have three reps: a proven closer (Dana), a solid mid-tenure rep (Marcus), and a six-week rookie (Tess). Your serviceable area scored out to 30 bricks totaling 900 opportunity points.

A naive even split would be 300 points each. But Tess cannot convert a 300-point prime territory the way Dana can — hand her one and her income craters while her map looks generous on paper. That is unfairness disguised as equality.

Here is a fair build using a blend:

Notice three things. First, the points are not equal — and that is the fair version, because skill is not equal. Second, Tess's lighter map is paired with a comp bump so her take-home is a fair shot, not a punishment. Third, effective selling hours are nearly identical, so nobody is secretly drowning in drive time.

Now a hailstorm hits Marcus's mixed-age area three weeks in. His 330-point territory jumps to an estimated 480 of real, time-sensitive opportunity. Left alone, Marcus cannot physically work it, good roofs go stale, and Dana and Tess watch from the sidelines. This is where the storm-override clause earns its keep.

Handling storms: the override that prevents the worst fights

Storms are where fair territory plans go to die, because a single event can triple one rep's opportunity overnight and leave the others on dead ground. If you have no rule for it, you get a land grab, and your team relationships pay for it. Write the rule before the storm, not during.

The storm-override clause

A workable version, in plain language you can put in your rep agreement:

"When a qualifying storm event materially increases the opportunity in one territory beyond what the assigned rep can work within [X] days, the storm-affected area is temporarily converted to a shared response zone. The originating rep retains a [Y]% override on all jobs closed by other reps in their territory during the response period, and full credit on everything they personally close."

That does three fair things at once: it gets the storm worked while the roofs are fresh and documentable, it does not punish the rep whose territory got hit (they earn an override on the help), and it does not let other reps feel locked out of the one event that matters all year. Set X around 7–14 days and Y around 10–20% based on your culture; what matters is that it is agreed in advance.

Working the storm without crossing the legal line

Storm season is also where roofing sales gets itself in legal trouble, and a fair, well-run team should not have reps freelancing claims promises to win doors. The clean, defensible role for your reps is documentation and an accurate estimate — nothing more.

Your reps may inspect the roof, photograph and document the damage thoroughly, and prepare an accurate repair estimate (Xactimate-aligned) for the work you would do. The homeowner is the one who files with their insurer, and the insurer decides coverage. That separation keeps you on the right side of the line.

What your reps must not say or do — teach this as a hard list and enforce it, because one cowboy on your team creates liability for the whole company:

• Do not negotiate, adjust, or "handle" the claim for the homeowner for a fee — that is unlicensed public adjusting in most states.
• Do not interpret the policy or tell the homeowner what is or isn't covered.
• Do not promise a specific payout, an approval, or that the claim "will go through."
• Do not promise the deductible will be waived, absorbed, eaten, or made to disappear. The deductible is the homeowner's legal responsibility; offering to erase it is insurance fraud in most states.
• Do not advertise or imply a "free roof."
• Do not represent the homeowner against the insurer.

The fair-territory angle here: if your comp plan secretly rewards reps for over-promising on claims, your most aggressive rep will dominate the storm zone by making promises that put your license at risk. Build the comp and the rules so the rep who documents thoroughly and estimates accurately wins, not the one who talks fastest about a payout. Your state's department of insurance defines the public-adjusting line; know it before storm season, not after a complaint.

Compensation and territory are the same problem

You cannot divide territory fairly if your comp plan fights it. The two have to be designed together. Here are the levers and how they interact with a fair split.

House accounts and the book problem

The biggest single fight in roofing sales territory is the book — past customers, referrals, and old estimates. If a tenured rep "owns" every past customer in their new territory, a transferred or new rep is handed dead, picked-over ground, and the tenured rep coasts on relationships they built years ago at a now-higher commission.

Fair handling, pick one and write it down:

• House accounts: the company owns past-customer and old-estimate follow-up centrally (often at a reduced commission since the relationship is company-built), and assigns those leads to whichever rep covers the territory now. This stops one rep from sitting on a goldmine.
• Origination credit with a sunset: the rep who originally closed a customer keeps referral/repeat credit for a set period (say 24 months), after which it reverts to the territory rep. Rewards the original work without permanent ownership of a street.
• Split on transfer: when a customer moves between reps, the originating and servicing reps split the next job. Simple, and it keeps both motivated.

Whatever you choose, the rule must exist before anyone needs it. Reverse-engineering a book rule during a dispute is how you lose a rep.

Self-generated vs. company-provided

A fair plan usually pays a higher rate on self-generated work (a rep's own cold-knocked or referred deal) than on company-provided opportunities (a brick handed to them with warm history or a marketed lead). This makes a "weaker" raw territory genuinely fair: the rep on fresh ground earns the higher self-gen rate on what they build, while the rep on a warm book earns the standard rate on inherited relationships. The map can be uneven if the rate structure offsets it.

Territory performance floors and forfeiture

Fair to the company matters too. A rep sitting on a prime, high-score territory and underworking it is starving the business and blocking a hungrier rep. A fair plan includes a performance floor: a minimum production expectation per opportunity point or per quarter, with a written, non-surprise process for re-cutting an underworked territory and reassigning bricks. "Use it or lose it," stated up front, is fair. Yanking streets with no warning is not.

The mistakes that quietly wreck fair territory plans

These are the failures we see most. Each one looks small and compounds into turnover.

1. Splitting on geography or door count. The original sin. A roof that is due ignores your ZIP lines. Balance opportunity, not acreage.
2. Using year-built as a proxy for roof age. Re-roofs are invisible in county and Zillow data. A 1985 house redone in 2020 is the worst door on the block and your map says it is the best. Score current-roof age, as a range.
3. Trusting a county hail polygon to balance storm work. Two reps "in the storm" can have wildly different real opportunity. Model exposure closer to the roof or you will hand one rep dead shaded map and another the actual damage and call it even.
4. Setting it and freezing it. Fair in January is unfair by April. No re-cut cadence means every adjustment feels like a punishment.
5. Tiering the map but flattening the pay. Giving rookies weaker ground is fine if the comp offsets it. Same rate on unequal maps is structural underpayment, and they will leave.
6. No book rule until there is a dispute. The past-customer goldmine causes more fights than anything else. Decide house-account handling before it matters.
7. No storm-override clause. The one event that makes the year turns into a land grab that poisons the team. Write the override before the storm.
8. No transparency. The single cheapest fix. Reps accept uneven maps when they can see the math. They never accept a map with no math, even a fair one.
9. Letting the loudest rep negotiate their own lines. If territory goes to whoever complains hardest in the parking lot, you are training your team to complain. Decisions come from the scorecard, not the volume of grievance.
10. Ignoring effective selling hours. Balanced opportunity with lopsided drive time is not balanced. Two selling hours a day lost to commute is a 25% pay cut nobody put in writing.

A quarterly re-cut cadence that keeps it fair over time

Fairness is a moving target. Run a light, predictable re-cut so adjustments are routine instead of personal.

The 15-minute weekly check

Once a week, glance at three numbers per rep: appointments set per knocking day, close rate, and any storm activity. You are watching for a territory that has gone cold (worked out, or a competitor moved in) or hot (a storm, a new development aging in). No action most weeks — just awareness.

The quarterly re-cut

Once a quarter, re-score the bricks and rebalance. Because territories are built from small bricks, you move a brick or two between reps to re-balance — not redraw the whole map. Publish the new scorecard. Rules to make it land:

• Announce the cadence up front so a quarterly adjustment is expected, not a surprise attack.
• Move bricks, not whole zones, so changes are small and explainable.
• Show the before/after score so a rep losing a brick sees they were over-weighted and are still within the fair band.
• Honor origination credit on anything already in motion in a moved brick, so a rep does not lose a deal they started by losing the street.

Trigger-based re-cuts

Some events do not wait for the quarter:

• A qualifying storm triggers the override clause immediately.
• A rep leaves triggers reassignment of their bricks by score (and a decision on their open book per your house-account rule).
• A new hire gets bricks from the rotating/overflow pool first, then graduates to exclusive ground as they ramp.

Pre-assignment checklist

Before you tell a single rep their new map, you should be able to check every box:

• The full serviceable area is drawn by real drive-time economics, not aspiration.
• It is broken into small bricks (150–400 homes), not big zones.
• Every brick is scored the same way: due-roof age, per-roof storm exposure, workability, openness, book equity.
• Territories are assembled from bricks and balanced within ~10% on total score.
• Effective selling hours are checked and balanced, not opportunity alone.
• Skill differences are handled through comp/tiering, written down — not through pretending unequal reps can work equal maps.
• A house-account / book rule exists in writing before anyone needs it.
• A storm-override clause exists in the rep agreement before storm season.
• Origination credit and self-gen vs. company-provided rates are defined.
• A performance floor and a non-surprise re-cut process are written.
• The full scorecard is ready to publish to the whole team.
• You can say the fairness sentence out loud and back it with the numbers.

Scoring a brick in practice: a full walkthrough

The scorecard is only useful if you can actually fill it in without a spreadsheet wizard on staff. Here is how to score one brick end to end, with the judgment calls spelled out, so your sales manager can do thirty of them in an afternoon.

Take a real-feeling brick: "Cedar Hollow," 280 homes, platted in 2001, a tidy grid of two-story colonials on quarter-acre lots about 12 minutes from your shop. A hailstorm clipped the north third of it 14 months ago.

Due-roof density (out of 40). The homes were built in 2001, so an original roof is now ~24 years old — squarely due. But a chunk of the neighborhood already reroofed: some after the hail 14 months back, some on their own over the years. Say your read is that roughly 55% of the 280 homes still carry a roof in the due range. That is ~154 due roofs in a tight, easy-to-work grid. On a 40-point scale where a fresh dense neighborhood of all-due roofs would max out, Cedar Hollow earns about a 30. If you only had year-built data you would have scored this near 40 and badly overstated it, because you would have counted the 45% that already reroofed as prime. That single correction is worth more than any other refinement in the whole model.

Storm exposure (out of 25). A storm hit, but 14 months ago and only the north third. Fresh, clearly-documentable storm opportunity is largely worked out by now in the hit area, and the southern two-thirds never took the hail. So despite "there was a storm here," the live storm value today is modest — call it a 6. Right after the event it would have been a 22. Storm score decays fast, which is exactly why a static map made last spring is lying to you this spring.

Workability (out of 20). Twelve minutes out, dense grid, quarter-acre lots, sidewalks, no gates. This is about as workable as residential gets. Score it 18.

Openness (out of 10). You have done maybe eight jobs here over the years and a competitor ran a campaign after the hail. Partly warmed by your signs, partly contested. Score it 6.

Book equity (out of 5). Eight past jobs and a couple of stale estimates. Score it 2.

Cedar Hollow total: 30 + 6 + 18 + 6 + 2 = 62. A genuinely good brick, but for boring reasons (age and workability), not the storm everyone remembers. A rep who fought to get this brick because "that's where the storm was" would be fighting over a 6-point input, not the 62-point reality. The scorecard turns a war story into a number.

Do this for every brick. The scoring takes two or three minutes each once you have the roof-age read in front of you. The discipline that matters is scoring every brick the same way — the absolute numbers can be a little rough as long as the method is consistent, because fairness is about relative balance between reps, not perfect precision on any one brick.

A quick gut-check on weights

If your market almost never sees hail, drop storm exposure to 10 points and push those 15 points into due-roof density (make it /55). If you are a storm-chasing-light shop that lives off events, you might weight storm higher. The defaults above suit a broad residential reroof shop that also works storms when they come. Pick weights once, write them down, and do not quietly change them mid-season — moving the weights after assignment is just gerrymandering with extra steps, and your reps will smell it.

Mapping it: the tools and the data you actually need

You do not need expensive territory software to start, though it helps at scale. Here is the honest tooling ladder.

Free / starter: Google My Maps plus a spreadsheet. Draw bricks as polygons, keep the scorecard in a sheet keyed to brick names, color the map by territory. This is completely adequate for a three-to-six-rep shop and forces you to think instead of letting software think badly for you.

Mid: A CRM with mapping (many roofing CRMs and canvassing apps do territory polygons, rep assignment, and pin-drop disposition tracking). The value here is that knock dispositions feed back automatically, so your saturation and openness scores update from real field data instead of memory.

Add-on data layers, in order of impact:

• Roof-age estimate by address or neighborhood — the highest-impact layer and the one most shops are missing. Without it you are scoring due-roof density off year-built, which is the core error.
• Per-roof storm exposure — modeled hail/wind likelihood at the roof, not the county polygon. This is what makes your storm score honest and your override clause fair.
• Parcel / lot-size data (often from county GIS or Census TIGER/Line) — feeds workability and drive-time estimates.
• Your own CRM job and estimate history — feeds saturation and book equity. You already own this; most shops just never put it on the map.

The common failure is buying mapping software and feeding it the same bad inputs (ZIPs, year-built, county hail). Pretty polygons on top of wrong data is still wrong data. Fix the inputs first; the map software is the cheap part.

Onboarding a new rep into a fair system

Fairness is not only between your tenured reps — it includes how you bring someone new into the map without robbing the people already on it or burning out the new hire on dead ground.

The ramp pool

Keep a small set of bricks designated as the "ramp pool": forgiving, workable, decent-but-not-prime ground that you do not permanently assign to anyone. New hires work the pool for their first 60 to 90 days. This does two fair things. It gives the rookie a real shot at money on ground that won't eat them alive, and it does not require you to carve streets out of a tenured rep's territory the day a new person starts — which is the fastest way to make your veterans hate every new hire on sight.

Graduating into exclusive ground

When a ramped rep hits their numbers, they graduate to exclusive territory. Where does that territory come from? This is where the small-bricks discipline pays off again: at the next quarterly re-cut, you peel a few bricks off the over-weighted territories (the ones running above the fair band) and assemble the new rep's area. Because you are moving bricks, not whole zones, and because the donor territories were over the band anyway, nobody loses ground they had a fair claim to. Show the donors the before/after scores so they see they are still inside the fair band.

The honest conversation about ramp ground

Tell the rookie the truth on day one: the ramp pool is intentionally forgiving, the comp during ramp is structured to give them a fair shot, and the prime ground comes when they earn it. Reps respect a system they understand far more than they respect being handed a fat-looking map they cannot work and watching their first paycheck come in thin. Honesty about the ramp is part of the fairness, not a softening of it.

Multi-crew and commercial wrinkles

The brick model above is built for residential reroof and storm reps. Two common situations bend it.

Commercial and large-account reps. Commercial roofing is relationship- and bid-driven, not door-density-driven, so the brick scorecard does not map onto it cleanly. Handle commercial as a separate book with named-account assignment (this rep owns these property managers, GCs, and facility groups) rather than geographic bricks. The fairness lever there is account quality and quantity balance plus a clear rule for who owns an inbound RFP that touches two reps' named accounts. Do not try to force commercial into a residential territory map; you will make both unfair.

Production capacity as a constraint. A perfectly balanced sales map can still be unfair if your install capacity cannot keep up in one area. If a rep sells faster than your crews can build in their territory, their backlog balloons, homeowners cancel, and the rep's effective income drops even though their map was "good." When you re-cut, sanity-check that each territory's expected volume fits the crew capacity that serves it. Sometimes the fair move is balancing sales territory against install territory, rather than opportunity alone.

Where the roof-age and storm data comes from

Everything above hinges on one hard input: a real read on which roofs in each brick are actually due, by age and by storm, before you draw lines. Most owners do not have this. They split on year-built and county hail maps because that is what is free and visible, and both of those are exactly the inputs that make a map look fair while being unfair.

This is the gap RoofPredict was built for. It takes aerial imagery and weather history per home and gives you a roof-age estimate as a range (not an exact date — re-roofs are invisible to records, so honest data gives you a tight range, not false precision), plus storm exposure modeled on each roof rather than a county-wide polygon. A hail map shows where it hailed; per-roof modeling shows which roofs the storm likely wore out, scored house by house. For territory work, that means you can score a brick on its real supply of due roofs and its real storm exposure instead of guessing from the house's build year.

Used for division, that does two concrete things. First, it lets you balance bricks on the variable that actually predicts jobs — due-roof density and per-roof storm exposure — so your scorecard is honest and your reps can see it. Second, it enriches your own list: you can take the streets you already plan to work and rank them by which doors are worth a rep's time, then skip the new roofs that waste mail, gas, and payroll.

Honest limits, because a fair team runs on honest tools: roof age comes back as a range, not a guaranteed install date. Storm modeling gives you odds of meaningful exposure, not proof that a given roof is damaged — your rep still has to get on the roof and document what is actually there. It is not a lead-buying service and it does not replace your reps' eyes; it sharpens the targeting and the territory math you are already doing. Used that way, it turns "who got the good side of town" from a parking-lot argument into a number everyone can see.

Putting it together

Fair territory division is not a one-time map. It is a system: score the ground on what actually drives jobs (roof age and per-roof storm exposure, not ZIPs and year-built), build territories from small bricks so you can rebalance without a war, pair uneven skill with offsetting comp instead of pretending it away, write the book rule and the storm override before anyone needs them, and publish the scorecard so every rep can see the math behind their map. Do that and the fights mostly stop, because there is nothing left to argue about that you cannot answer with a number.

The reps who quit over "a bad territory" were usually right that it was bad. They were wrong that it was personal. Make the math visible and the badness fixable, and you keep the good ones.

If you want to balance territory on the one input most maps get wrong — which roofs are genuinely due, by age and by the storms they have actually taken — that is exactly the read RoofPredict gives you, house by house, before anyone gets on a ladder. Book a demo and bring a street you already know well; you decide whether the roof-age ranges and storm scores match what your crew has seen on the ground.

FAQ

Should I split roofing territory by ZIP code or square miles?

No. ZIP codes and acreage are the easiest way to split and the least fair, because a roof that is due ignores your boundary lines. Balance expected opportunity per selling hour instead: due-roof age density, per-roof storm exposure, drive time, and saturation. Two reps can have identical square miles and wildly different income potential.

How do I make territory fair when my reps have very different skill levels?

Do not force equal maps onto unequal reps. Give stronger closers the higher-opportunity ground and lighter ground to ramping reps, then offset the difference through compensation — a higher commission rate or a ramp guarantee on the weaker territory. Fairness comes from equal take-home opportunity, not identical maps. The catch: the tiering and the comp offset must be written down and open, because hidden tiering is the most resented thing in roofing sales.

What is the fairest way to handle storm leads across territories?

Use a storm-override clause written before storm season. When a qualifying event overloads one territory beyond what the assigned rep can work in 7 to 14 days, convert it temporarily to a shared response zone. The originating rep keeps a 10 to 20 percent override on jobs other reps close in their area plus full credit on what they personally close. That gets the roofs worked while they are fresh, does not punish the rep whose ground got hit, and does not lock everyone else out of the year's biggest event.

Why can't I just use year-built data to find the old roofs?

Because year-built is the age of the house, not the age of the current roof. Re-roofs are invisible in county records and on Zillow, so a 1985 house redone after a 2020 storm shows as ancient and is actually the worst door on the block. You need an estimate of the current roof's age, ideally as a range, neighborhood by neighborhood, to score territory honestly.

How often should I re-cut roofing sales territory?

Run a light quarterly re-cut as a published, expected cadence, plus a quick weekly glance at appointments-per-day and close rate per rep to catch a territory going cold or hot. Some events trigger an immediate re-cut: a qualifying storm fires the override clause, a rep leaving triggers reassignment by score, and a new hire pulls from a rotating overflow pool first. Because territories are built from small bricks, you move a brick or two to rebalance rather than redrawing the whole map.

How do I handle past customers and old estimates when a rep takes over a territory?

Decide the book rule in writing before anyone needs it, or it becomes the biggest fight on your team. Common fair options: make past customers and old estimates house accounts the company owns and assigns to the current territory rep at a reduced rate; give the original rep origination credit that sunsets after 18 to 24 months; or split the next job between the originating and servicing reps on transfer. The wrong move is letting one tenured rep permanently own every past customer on a street while a new rep is handed picked-over ground.

What size should I make my territory building blocks?

Use small bricks of roughly 150 to 400 homes — a single subdivision or a census-block-group-sized chunk. Small bricks are what make ongoing fairness possible: you assemble territories out of them and can shift one brick between reps later to rebalance without redrawing the entire map or starting a turf war.

Can my reps mention insurance claims when working a storm territory without legal trouble?

Reps can inspect the roof, document damage thoroughly with photos, and prepare an accurate repair estimate for the work you would do — then hand it to the homeowner, who files and whose insurer decides coverage. They must not negotiate or handle the claim for a fee, interpret the policy or what is covered, promise a payout or approval, promise the deductible is waived or absorbed, advertise a free roof, or represent the homeowner against the insurer — those cross into unlicensed public adjusting or fraud in most states. Teach that list as a hard rule and build comp so the rep who documents thoroughly wins, not the one who over-promises.

How do I keep a rep from sitting on a great territory and underworking it?

Build a written performance floor into the plan: a minimum production expectation per opportunity point or per quarter, with a clear, non-surprise process for re-cutting an underworked territory and reassigning bricks to a hungrier rep. Stated up front, use-it-or-lose-it is fair. Yanking streets with no warning is not, and it will cost you trust with the rest of the team.

What's the single most important thing for making reps accept their territory as fair?

Transparency. Publish a scorecard showing each territory's total opportunity score and the breakdown — due-roof density, storm exposure, workability, openness, book equity — alongside everyone else's. Reps will accept an uneven map when they can see the math behind it. They will never accept a map with no math, even a perfectly fair one, because they assume the worst about anything hidden.