About 30 doors per signed contract if you're average. About 11 if you're good.
That's the short answer, and it comes from our 2026 State of Storm Restoration survey — 247 contractors across 31 states, with their reported funnel numbers cross-checked against tracked canvassing data.
The longer answer is more useful, because the single ratio hides where the money actually leaks. A rep isn't losing 29 out of 30 doors evenly across the funnel. They're losing almost all of them at the first step, and once you can see that, canvassing stops being a grind and becomes something you can coach.
The full funnel, stage by stage
| Funnel stage | Median | Top quartile |
|---|---|---|
| Doors knocked → appointment set | 9% | 16% |
| Appointment → inspection completed | 71% | 84% |
| Inspection → signed contract | 58% | 74% |
| Contract → completed install | 88% | 95% |
Run 1,000 doors through both funnels and the compounding becomes obvious:
| Stage | Median rep | Top-quartile rep |
|---|---|---|
| Doors knocked | 1,000 | 1,000 |
| Appointments set | 90 | 160 |
| Inspections completed | 64 | 134 |
| Signed contracts | 37 | 99 |
| Completed installs | 33 | 94 |
Same 1,000 doors. 33 installs versus 94. Nearly three times the production from identical effort, and no single stage required a heroic improvement — each one got modestly better and the gains multiplied.
That's the number worth internalizing: a signed job every ~27 doors at the median, every ~10 doors in the top quartile. Add the jobs that fall out before install and the contract math is one per 30 versus one per 11.
Where the leak actually is
Look at the gap at each stage in absolute terms, not percentage points.
| Stage | Extra deals a top-quartile rep gains per 1,000 doors |
|---|---|
| Doors → appointments | +70 appointments |
| Appointments → inspections | +25 inspections (of which ~15 trace back to the door step) |
| Inspections → contracts | +18 contracts (of which ~13 trace back upstream) |
| Contracts → installs | +5 installs |
The door-to-appointment step is where nearly all of it comes from. A rep who improves their close rate from 58% to 74% while still converting 9% of doors gains about 10 contracts per 1,000 doors. A rep who improves their door conversion from 9% to 16% while closing at the median gains about 29. The first step is worth roughly three times the last one, and it's the cheapest to coach.
This is why the shops that grow fastest obsess over leading indicators. A rep can't directly control installs. They can control doors knocked and appointments set today, and everything downstream follows.
What a realistic day looks like
Knock counts get exaggerated in this industry, usually by people selling training. Here is what actually holds up:
| Metric | Realistic range |
|---|---|
| Doors per hour | 15–25 |
| Productive canvassing block | 4–5 hours |
| Doors per full day | 60–100 |
| Doors per month (20 days) | 1,200–2,000 |
| Signed contracts per month, median | 44–74 |
Those hourly numbers assume real work: walking the street, having actual conversations at the doors that open, and logging the outcome before moving on. A rep claiming 200 doors a day is either counting houses they walked past, or not having conversations at the ones that answer.
The contact rate is the hidden variable. Only 25–40% of doors open at all, depending on time of day and neighborhood. Everything in the 9% figure is already net of that — so a 9% doors-to-appointment rate is closer to a 25–35% contact-to-appointment rate, which is a much more encouraging way to look at your own performance.
The variable that beats technique: was there actually a storm?
All of the numbers above assume a neighborhood with verified storm damage. Take the same rep and the same script to a neighborhood that hasn't been hit and the ratio can degrade to several hundred doors per job.
That's not a scripting problem. It's a targeting problem, and it's the reason storm restoration canvassing lives or dies on knowing which streets actually took hail before anybody walks them. A rep working a confirmed swath at 9% conversion will out-produce a better rep working clean roofs at 1%, every single time.
Two practical consequences:
- Verify the damage before you build the route. Radar-derived hail swaths tell you which streets took what size hail on what date — that date being the date of loss the whole claim will hang off. Our guide on how to read a hail map covers what the colors actually mean.
- Knock the fresh swath first. The first seven days after a confirmed hail event are worth more than the next thirty combined, because you're competing against fewer roofers and a homeowner who hasn't already signed with one.
How to measure your own ratio honestly
Your number is only as good as your denominator, and this is where most companies quietly fool themselves.
If reps log the doors that went well and skip the ones that didn't, your knock-to-contract rate looks fantastic and means nothing. If the denominator is a tally sheet filled in from memory at the end of the day, it's a record of what people remember, not what happened.
To get a number you can actually manage against:
- Log every door on the map, including "not home" and "not interested". A canvassing pin with a timestamp and a GPS location is a fact; a tally mark is a claim.
- Break the ratio into stages. Doors→contacts is a timing and territory problem. Contacts→appointments is a script problem. Appointments→inspections is a follow-up problem. Inspections→signed is a damage-quality and closing problem. A single blended number hides which one is broken.
- Compare reps on the same neighborhoods. A rep working fresh swaths will beat a rep working week-old ones regardless of skill. Rank on the same territory or you're measuring luck.
- Track it weekly, not monthly. A storm season is short. A metric you review in October can't fix a July that went wrong.
The mechanics of that — what to track and how to coach on it without micromanaging — are covered in Rep Performance Metrics Every Roofing Manager Should Track.
Turning the ratio into a target
Work backwards from revenue instead of forwards from effort.
Say you want 10 signed jobs a month from one rep, at median conversion:
- 10 contracts ÷ 58% inspection close = 18 inspections
- 18 inspections ÷ 71% show rate = 25 appointments
- 25 appointments ÷ 9% door conversion = 278 doors
278 doors a month is about 14 a day over 20 working days — well under a full canvassing block. Which means at median rates, one rep knocking a genuinely full day should be signing far more than 10 jobs a month, and if they aren't, one of two things is true: they aren't knocking as many doors as the count says, or the neighborhoods aren't hit.
Run that same arithmetic with your own conversion rates and you get a daily door target that ties directly to the revenue number on the wall. That's a target a rep can act on at 8am. "Sell more roofs" isn't.
The bottom line
Thirty doors per job at the median, eleven in the top quartile — and the entire gap is concentrated in the first ten seconds at the door. Improve that one step and every downstream number moves with it.
But the biggest single lever isn't at the door at all. It's picking the street. A verified hail swath, worked in the first week, with every door logged so nobody knocks it twice, will beat better scripting on the wrong block every time.
That's the whole idea behind HailMate's canvassing tools: the hail map picks the streets, the pins keep the count honest, and the ratio you end up managing is a real one.