management

Hiring and Ramping Door-to-Door Canvassers

What the labor data says about the canvasser role, how to screen for it, and why the classification question comes first.

Hiring canvassers is not hard. Keeping them is hard, and most of the difficulty is created in the first two weeks by decisions the hiring manager made before the rep showed up.

Here is what the actual data says about the role, and what to do with it.

What the job pays, according to the government

The Bureau of Labor Statistics tracks "Door-to-Door Sales Workers, News and Street Vendors, and Related Workers" as occupation code 41-9091. There is no full Occupational Outlook Handbook profile for it, so the figures live on the BLS page for occupations not covered in detail: a May 2025 median annual wage of $41,380, 2025 employment of 50,900, and a projected change of negative 5,300 jobs, or negative 10 percent, over 2025 to 2035.

Two things about that figure.

First, it is a median across a broad category that includes street vendors and news sellers, not a roofing or solar canvasser specifically. It is the floor of the conversation, not the market rate for a good hire in a hail market.

Second, the projected decline matters for how you position the job. If you recruit against the category, you are recruiting into a shrinking occupation. If you recruit as a route into field sales, you are recruiting into "Sales Representatives of Services," code 41-3091, which the same BLS page lists at a May 2025 median of $69,990 with 1,272,700 employed and a projected increase of 33,900 jobs.

That reframe is honest and it is the one that attracts people worth keeping. Canvassing is the entry point to a field sales career, and the people who last are the ones who understood that when they took the job.

Settle the classification question before you make an offer

This is the part most teams get wrong, and it is expensive.

Two separate questions, often confused.

Is the rep an employee or an independent contractor? The Department of Labor's Fact Sheet #13 is direct about what does not decide this: "What the worker is called is not relevant," signing an independent contractor agreement "does not make a worker an independent contractor under the FLSA," and a 1099 does not either. The analysis runs on economic-reality factors including control, opportunity for profit or loss, investment, permanence, skill, and whether the work is integral to the business.

The regulatory ground here has been moving. DOL's Field Assistance Bulletin 2025-1 states that the Wage and Hour Division "will no longer apply the 2024 Rule's analysis when determining employee versus independent contractor status in FLSA investigations," while noting "the 2024 Rule remains in effect for purposes of private litigation." A 2026 rulemaking proposes to rescind the 2024 rule and center the analysis on control and opportunity for profit or loss. Check the current status before you rely on any of it.

Apply it to a canvasser honestly. Someone who works an assigned territory on your schedule, with your script, your leads, and no capital investment, is weak on exactly the factors that matter most.

If they are an employee, are they exempt? The outside sales exemption is where roofing and solar teams get surprised. DOL's Fact Sheet #17F sets a two-part test: the primary duty must be making sales, and the employee must be "customarily and regularly engaged away from the employer's place or places of business." The exemption carries no salary floor.

The regulation at 29 CFR 541.502 says outside sales happen "at the customer's place of business or, if selling door-to-door, at the customer's home," so a true door-knocker clears the location prong. But 29 CFR 541.503 is the trap: "promotional work that is incidental to sales made, or to be made, by someone else is not exempt outside sales work."

A canvasser whose job is to set appointments for a closer is, on the face of the regulation, doing promotional work for somebody else's sale. That is the classic non-exempt pattern, which means minimum wage and overtime apply. A setter-and-closer model is extremely common in roofing and solar and it is the model most likely to fail this test. Get an employment lawyer to look at your specific structure before you build a comp plan on top of it.

Screen for three things and ignore the rest

Canvassing selects on a narrow set of traits, and sales resumes predict almost none of them.

Tolerance for repetitive rejection. This is the job. Not resilience in the abstract, the specific ability to take forty consecutive noes and knock the forty-first door with the same energy. The best interview question is a behavioral one about a time they did something repetitive and unrewarding for a long stretch, and the best answer is boring rather than heroic.

Willingness to be outside. Heat, cold, rain, hills, dogs, six miles of walking. People self-select out of this fast and it is better for everyone if they self-select out before training. Say the physical reality plainly in the job post.

Coachability on the first fifteen seconds. The door-opening is the one teachable skill that moves the numbers most. A candidate who takes a correction in a role play and applies it immediately is worth more than a candidate with a better natural pitch who argues.

What does not predict performance: prior industry experience, polish, and the ability to deliver a smooth monologue in an interview room. Interviews reward verbal fluency. Doors do not.

Run a ride-along as the final stage. Half a day in the field with a veteran filters more effectively than any interview, and it is honest in both directions.

The ramp is about three things, and they are measurable

The best peer-reviewed evidence on new-hire ramp is a meta-analysis by Bauer, Bodner, Erdogan, Truxillo, and Tucker, Newcomer adjustment during organizational socialization, published in the Journal of Applied Psychology in 2007 and drawing on "70 unique samples of newcomers." It found that three adjustment variables, "role clarity, self-efficacy, and social acceptance," mediated the effects of onboarding on job satisfaction, commitment, performance, intentions to remain, and turnover.

Translate those into canvassing.

Role clarity. Does the new rep know exactly what a good day looks like? Not "sell hard," but a door count, a territory, an hour window, and a definition of each disposition. Ambiguity here is the leading cause of a week-two quit, because a rep with no target cannot tell whether they are failing.

Self-efficacy. Do they believe they can do it? This is why the first week should be engineered to produce a success, even a small one, rather than to test them. Put a new rep in a territory with known decent numbers, not in the worst turf because nobody else wants it.

Social acceptance. Have the veterans accepted them? A new rep who eats lunch alone in week one usually does not have a week four. Pair them, ride them along, and put their name on the board early.

Notice what is not on that list: product knowledge. A canvasser needs enough to be truthful and enough to hand off. Front-loading four days of shingle specifications buys role clarity you could have bought in an hour and crowds out the thing that actually matters, which is repetitions at real doors.

Expect short tenure and plan for it

BLS reported in its January 2026 Employee Tenure release that median tenure for all wage and salary workers was 4.1 years, 3.0 years for ages 25 to 34, and that 74.1 percent of 16 to 19 year olds had tenure of 12 months or less.

Canvassing recruits heavily from the young end of that distribution. The implication is not that you should accept churn. It is that your training has to be repeatable by design, because you will run it often, and your territory data has to live in the system rather than in a rep's head, because the rep will leave and the territory will not.

A team whose callbacks live in a departing rep's phone does not just lose the rep. It loses the street.

Canvass Pro keeps the territory, the door history, and the callbacks on the territory record rather than on the person, so a rep leaving does not take the neighborhood with them.