Guide 06

Field Team Management: Hiring, Paying, and Keeping Canvassers

Every canvassing operation is a retention operation wearing a recruiting operation's clothes. Programs measure how many people they hired and how many doors got knocked, and almost none of them measure the thing that determines both: how many people were still there in week four.

That number drives everything. A canvasser in their fourth week works at close to double the rate of a first-week canvasser, records better data, and can be handed the difficult turf. Losing them and hiring a replacement costs the full training cycle again, plus the territory they knocked badly on the way out. This guide is about the practices that keep them, in the order you need them.

The week-three problem

Canvassing attrition is front-loaded to an unusual degree. The people you lose, you mostly lose before they were any good, and the mechanism is not mysterious.

A new canvasser spends their first two weeks being told no by strangers, at a rate they were not warned about, while working at roughly half the speed of the person who trained them. Under commission-only pay they are also earning very little during exactly that window. Nothing about the experience signals that it gets better, because from inside it does not look like it is getting better.

Everything in this guide that works, works by shortening the distance to the first moment a canvasser feels competent. That is the whole strategy. Better territory for new people, pairing, daily coaching, telling them the real contact rate in advance so seven not-homes in a row reads as a normal afternoon rather than personal failure.

The programs that treat the channel as a filter, hiring in large classes and letting attrition sort them, are paying a cost they cannot see: every person who quits leaves behind turf that was knocked badly and cannot be re-knocked soon, which makes the territory worse for whoever comes next.

Where good canvassers come from

Recruiting channels for this job sort very differently from how they sort for office roles, because the trait that matters is not on a resume.

What predicts survival is prior experience of being turned down in person, repeatedly, without taking it personally. Restaurant and bar work, retail, coaching, teaching, other field sales, campaign work, and organizing all produce people who have already met that. Someone whose entire work history is screen-based is not disqualified, but they have not been tested against the part of the job that removes people.

The highest-yield channel is almost always referral from a canvasser who is doing well. They understand the job, they only refer people they think can do it, and the referred person arrives with a realistic picture and a friend on the team. Both of those show up directly in retention. Referral bonuses paid on the referee reaching a tenure milestone, rather than on hire, align the incentive correctly.

Job boards produce volume and poor retention, which is fine if you know that going in and are staffing a spike. What they do not produce is a stable core team.

For volunteer programs the dynamics are different again and the single strongest predictor is a direct personal ask from someone the volunteer already knows. Broadcast recruitment produces sign-ups; personal asks produce attendance, and those are not the same number, as the political canvassing guide covers in detail.

Interviewing for the job that actually exists

Most canvasser interviews test whether someone is pleasant in a conversation they are comfortable in, which is not the skill.

Two things are worth doing instead. First, describe the job honestly and watch the reaction. Say the contact rate out loud. Say that most doors do not answer, that some people are rude, that the first two weeks are the worst two weeks. A candidate who leans in after hearing that is telling you something real, and one who visibly deflates has done you both a favor now rather than in week two.

Second, make them do it. A five-minute role play at a pretend door tells you more than half an hour of questions. You are not looking for polish. You are looking for whether they keep going after the first brush-off, whether they can hear what the person said and respond to it, and whether they can be coached inside the exercise: give one correction, run it again, see if it takes. Coachability is the trait that separates people who improve from people who plateau, and it is directly observable in ten minutes.

Be wary of over-indexing on charisma. The consistently high performers in field work are more often methodical than magnetic. The person who knocks their whole list in order and records everything beats the charmer who works four doors brilliantly and forgets the rest.

Pay structures and what each one rewards

Whatever you pay for is what you will get, and this matters more in field work than almost anywhere else, because the canvasser is alone all day making small unobserved decisions.

Per door produces doors. It also produces short conversations, skipped not-home records, and a strong pull toward the easy streets. Since conversation quality is the thing that actually moves outcomes, this structure optimizes against your own goal.

Per outcome, meaning per ID, per signature, per appointment, produces outcomes, including ones that did not happen. Any per-outcome structure needs verification attached to it or you are paying for a number rather than a result. This is covered in the canvassing operations guide under quality control, and it is not a hypothetical risk.

Hourly removes the perverse incentives and introduces a different one: nothing rewards pace. It works when supervision is close and coaching is real, and it drifts when it is not.

Base plus variable is what most teams that have solved retention end up with, at least for the first several weeks. The base is not generosity, it buys the ability to direct behavior. You can tell someone to work the hard turf, spend longer qualifying, or take a trainee out, and the instruction is credible because their rent does not depend on ignoring it.

If you use a draw against commission, cap it and watch the balances. A rep deep in a negative draw is working while owing money, which is the shortest path to both a resignation and falsified paperwork.

One more, because it is where teams quietly go wrong: be careful what the leaderboard ranks. Ranking on raw closes rewards whoever drew the best territory and punishes whoever drew the worst, which drives out people who were doing fine. Rank on something territory-adjusted, and never on a metric that improves when someone stops recording not-homes.

The first shift

The first shift decides a large share of your retention, and it is cheap to run well.

• Role play before they go out. Twenty minutes, including one deliberately difficult door. Reading a script silently is nothing like saying it to a person, and the first time should not be on a stranger's porch.
• Pair them. Let them watch three or four doors before they take one.
• Give them good turf. The instinct to hand new people the leftovers is exactly backwards. They need an early win more than your best rep does.
• Tell them the numbers first. If they know the contact rate, a bad stretch reads as normal. If they do not, it reads as evidence they cannot do this.
• Debrief the same night, in person. Ask what surprised them. This is where you catch the thing that would have made them quit.

Training detail beyond the first shift is covered in the complete guide to door-to-door canvassing.

Coaching, and the ride-along

The highest-return management activity available to you is standing on a sidewalk watching somebody work, and it is the first thing that gets cut when the week gets busy.

Numbers tell you a canvasser is not converting. They do not tell you whether the opener is landing fine and the whole thing collapses at the transition, which is a twenty-minute fix, or whether they are knocking eleven doors an hour and calling it a day, which is a completely different conversation. Only watching tells you which.

A ride-along that works has a shape. Watch several doors saying nothing, because a person being coached mid-door performs rather than works. Pick one thing. Say it as a specific action for the next door rather than a general quality. Then watch a few more and see whether it took. A canvasser given five corrections applies none of them.

Daily beats weekly by a wide margin in the first fortnight. Ten minutes at the end of a shift about the two doors that went worst compounds. A Friday review of a week nobody remembers in detail does not.

Watch for one thing that never appears in the numbers: people who have quietly stopped knocking the doors they expect to fail. It feels like efficiency to them, and it is how a territory gets burned without anyone deciding to burn it.

Performance conversations

Two failure modes, and most managers have one of them. Some avoid the conversation until they are firing someone. Others deliver a running commentary that nobody can act on.

What works is narrow and specific. Bring one metric and one observation, not a dashboard. Separate the things the canvasser controls from the things they do not, out loud, because a person being held responsible for their territory's quality will disengage and they will be right to. Agree on one change and a date to look at it again.

Before any of that, check the obvious: is this person's territory comparable to the people you are implicitly comparing them to? A rural canvasser at nine doors an hour may be outperforming a suburban one at twenty-two. If your reporting cannot separate those, fix the reporting before you have the conversation, or you will lose good people for the crime of being assigned bad turf.

Volunteers are a different problem

Volunteers cannot be managed with any of the levers above. There is no pay to adjust and no consequence for not showing up, which changes the whole approach.

What retains volunteers is being useful, being known, and seeing the result. Concretely: give them the same turf each time so familiarity builds, confirm the day before rather than trusting a sign-up from three weeks ago, put them in pairs, have somebody who knows their name greet them at launch, and tell them afterward what the data showed. The last one is skipped constantly and it is the one that makes people feel the shift mattered.

Plan for the gap between commitments and attendance rather than being surprised by it every Saturday. The arithmetic for that, and how it feeds coverage planning, is in the political canvassing guide.

Scaling for a seasonal spike

Field programs are spiky by nature. A campaign triples headcount for six weeks. A roofing team doubles after a storm. Pest control ramps every spring. Hiring for a spike is a different exercise from hiring for a core team, and the mistake is running the second playbook at the first problem.

Three things to get right. Start earlier than feels necessary, because anything with a background check or a per-canvasser permit has a lead time that is invisible until it blocks you; the permit side is in the canvassing compliance guide. Protect your tenured people, because a spike dilutes supervision and that is exactly when your good canvassers get ignored and start looking around. And work out what the spike costs you in software and overhead before you commit, since per-seat annual pricing punishes precisely this shape of business.

What to actually measure

Most field programs measure hiring and door counts, which together tell you almost nothing about whether the team is healthy. Four numbers that do:

• Week-four survival rate. The share of new hires still working after four weeks. This is the headline number and almost nobody tracks it.
• Time to first outcome. How long until a new canvasser gets their first sale, ID, or signature. Everything that improves retention shortens this.
• Productivity by tenure. Attempts and conversion split by weeks on the job, so you know what your ramp actually looks like instead of guessing.
• Referral share of hires. A proxy for whether the people already doing the job would recommend it, which is the most honest signal you have.

None of these are visible if reporting only rolls up to team totals, and none are meaningful if you cannot separate territory quality from canvasser performance.

Common questions

Why is canvasser turnover so high?

The first two weeks are almost entirely rejection, and under commission-only pay a new canvasser is losing money during exactly the period when they are worst at the job. Most attrition happens before anyone has had time to get good.

Where do you find good canvassers?

Referrals from canvassers who are doing well, and people whose previous work already involved being turned down in person. Job boards produce volume and poor retention.

What should you pay canvassers?

Decide what behavior you want first, because whatever you pay for is what you will get. Most teams that have solved turnover include a base for at least the first several weeks, since it buys the ability to direct behavior.

How long before a new canvasser is productive?

A first-week canvasser works at roughly half the rate of an experienced one. Most people approach normal productivity in the third or fourth week, which is why losing them in week three is so expensive.

How many canvassers does one manager need?

It depends on tenure, since new crews need far more supervision. The practical limit is that a manager who cannot ride along with each person regularly has too many people.

Where to go next

The doorstep craft your team is being trained in is in the complete guide to door-to-door canvassing. Territory quality, which decides how fair your performance numbers are, is in the canvassing operations guide. Hiring lead times created by permits and background checks are in the canvassing compliance guide.

On the software side, most of this guide depends on being able to tell territory quality apart from canvasser performance. That is what Beacon's analytics and reporting is built to do: results by territory and by tenure, not just a leaderboard.