Guide 04

Canvassing Operations: Territory, Routing, and Measurement

When a canvassing program (an organized effort to knock on doors) underperforms, the problem is usually not at the door. It is in how the work is planned and managed. Turf (the area assigned to each canvasser) is divided badly, walk lists (the addresses each canvasser visits) are in an order that wastes an hour a shift, a manager learns on Tuesday that Saturday went wrong, or everyone is chasing a number that can rise without the program improving.

This guide covers that management work: how to divide and adjust turf, why the order of addresses matters, what a field director does during a shift, how to choose software without buying a CRM (customer relationship management system) you will not use, and how to measure a canvass so that a rising number means the program is improving. Talking with people at the door is covered in the complete guide to door-to-door canvassing.

Cutting turf

Cutting turf means dividing your target area into territories, and a territory should be the work one canvasser can do in one shift. Planning around that definition rules out the most common method, which is dividing by equal record counts.

Two hundred records in a dense grid of townhouses and two hundred records spread along county roads are very different amounts of work. Divide turf by estimated walking time first, then look at the record counts and accept that they will be uneven. If every territory in a program has the same number of doors, the turf was almost certainly divided incorrectly.

The second rule is that boundaries should follow real features, such as streets, rivers, railway lines, and precinct edges. A boundary drawn across the middle of a block leaves canvassers unsure whether the house in front of them is theirs, and it leads to doors being knocked twice where two territories overlap.

Political programs have one more rule worth following, even when it makes the boundaries less tidy: keep turf inside precinct lines. Election results are reported by precinct, so territories that cross precinct boundaries can never be compared against results afterwards, and you lose the ability to learn anything from the election.

Rebalancing, and why it never happens

Turf is usually divided once, at the start, using assumptions that turn out to be wrong within two weeks. One territory is much slower than expected, another contains an apartment complex nobody planned for, and a third turns out to be half rental homes. Everyone can see these problems by week three, and almost nobody fixes them.

The reason is that changing turf is often difficult. If redrawing turf means a support request, a data vendor, or an evening of rebuilding paper packets, it will not happen, and the program will run a whole season on territories known to be wrong in week two. The practical lesson is that being able to redraw a boundary in a few minutes is worth more than any amount of care in the original plan, because the original plan is always going to be wrong.

When you do rebalance, move the boundaries rather than reassigning people. Canvassers get faster on turf they know, because they learn which houses have dogs, which blocks are home in the evening, and where the shortcuts are. Moving people to new territories loses that knowledge.

Route order

Within a territory, the order of the addresses on the walk list has a real effect on cost, and fixing it costs nothing.

An unordered list, or one sorted by name or by ID, sends a canvasser crossing and re-crossing the same streets. A list ordered so the canvasser works up one side of a street and down the other removes that entirely. The gain is several attempts (knocks at an address) per hour. Multiplied across every canvasser and every shift in a season, that equals an extra person you did not have to recruit, train, or retain.

Ordering has a second benefit that people often miss: it makes coverage easy to check. When a canvasser works a street in order, a manager can look at the recorded dispositions (the result entered for each door) and tell immediately whether the street was actually walked. When results are scattered across a territory, it is much harder to tell whether the canvasser chose only the easiest doors.

Running the shift

A shift has six stages: launch, dispatch, walk, check-in, return, and debrief. The two that get dropped when things are busy are the launch and the debrief, and those are the two that matter most.

The launch is the short meeting before canvassers head out, where the day's goal is stated, the script is reviewed, and anyone who is clearly not ready is noticed before they are alone on a street. The debrief, the meeting at the end, is where the data comes back, where the script is corrected based on what actually happened, and where people decide whether they will return. A program that only sends people out and collects results, with nothing at either end, loses both data quality and people.

During the walk, the field director watches for three things, and each can be acted on only while the shift is still running. The first is whether anyone has stalled, which usually means they are stuck, lost, or having a bad time. The second is whether any territory is far off pace, which usually means it was drawn badly. The third is whether dispositions are being recorded at all. If nobody checks this one, missing records can ruin a program unnoticed.

This is the practical reason to use live data instead of data that updates only at the end of the night. Learning on Tuesday that Saturday went badly only tells you what happened. Learning at 4pm that one canvasser has recorded nothing since 2pm still leaves time to act.

How many passes, and when

A single pass (one round of visits) through a universe (the full set of people or households you are trying to reach) reaches only a minority of it. That is not a failure of the canvass. Most people are not home most of the time, and the people who are home on a Saturday afternoon are not typical of the neighborhood as a whole.

Planning for several passes is therefore essential. Without it, you cover mostly the people who happen to be retired rather than the whole universe. These are the rules that matter:

• Vary the time of day between passes. Knocking on the same street at the same hour three times mostly reaches the same households each time and never reaches the rest. At a minimum, alternate between weekday evenings and weekend daytime.
• Carry not-homes forward, and drop completed addresses. The second pass should mostly cover the addresses the first pass missed, rather than a fresh walk of the whole territory. After route order, this is the largest efficiency gain available, and it is the one that paper-based programs cannot do.
• Stop at three attempts for most programs. The fourth attempt at an address that has never answered is worth less than a first attempt somewhere new. The exception is when a deadline makes coverage of a specific universe the only thing that matters.
• Space the passes apart. Two passes in the same week feel like pestering to a household, and they reach nearly the same people.

As a result, the number of shifts you plan has to be based on passes, not on universe size alone. A 12,000-record universe at two and a half passes is 30,000 attempts, and campaigns that plan for 12,000 discover the gap far too late to fix it.

Data discipline

Three rules cover most good record-keeping.

Record every attempt, not only every contact. Addresses where nobody was home are useful information. They tell you which streets need an evening pass, and you need them to calculate your contact rate (the share of attempts that lead to a conversation). A program that only records conversations has a contact rate of 100 percent and no idea how it is doing.

Record at the door. Anything written from memory at the end of the night is partly guessed, especially the details. For this reason, offline capture (recording results with no cell signal) is a requirement for accurate data, not an optional feature. Without it, a canvasser in a basement corridor or on a rural road with no signal will write the result down later, or not at all.

Keep the record on the address, not the person walking it. Canvassers and sales reps leave and are replaced constantly. If the callback, the note about the dog, and the fact that the homeowner works nights exist only in one person's memory or paper packet, all of it is lost when that person leaves. This is the largest unnoticed loss in programs with high turnover.

Quality control

Every canvassing program eventually has to ask an uncomfortable question: is the data accurate? Canvassers are rarely dishonest, but the incentives point one way and supervision is limited. A person alone on a street, measured on doors, with a long walk to the next house, tends to keep optimistic records even when acting in good faith.

The two kinds of problem look different. Skipping is a canvasser quietly not knocking on doors they expect to go badly, then recording them as not-home. It feels efficient, but it ruins your not-home list, which the second pass is built from. Fabrication is recording conversations that did not happen. It is rarer, and much worse, because a fabricated ID (a recorded answer about where someone stands) goes into the universe you build your final week on.

Three checks catch nearly all of it, and none of them require treating anyone as a suspect.

• Look at where the results were recorded. A canvasser who walked a street in order leaves an orderly trail on the map. Records scattered across a territory, or a group of results recorded from one spot, are worth a conversation.
• Look at the timing. Records entered all at once at the end of a shift were written from memory, which at best means the details are approximate.
• Call back a sample. Call five reported contacts a week and ask whether someone came by. This is standard practice in paid canvassing operations, and it is the only check that actually confirms a conversation happened.

The approach that works is to make the checks routine and openly known, rather than starting an investigation. Canvassers who know their results can be seen on a map record honestly, and the few who were skipping doors stop. Announcing an audit after you already suspect someone puts you in a much worse position.

One caution: surprisingly good numbers deserve the same attention as surprisingly bad ones. Programs usually investigate underperformers and celebrate top performers, which is the wrong approach if what you care about is whether the data is accurate.

Choosing software

Canvassing software falls into three categories that are often confused with one another when teams compare products.

• Data platforms hold the master voter or customer file. They are strong on targeting and matching records, usually weaker as a phone app, and often priced and restricted by which organizations are eligible.
• Field sales trackers focus on the individual salesperson. They are strong on tracking sales in progress, quotas, and management reports, with canvassing as an add-on rather than the center of the product.
• Field operations tools focus on the walk itself. They are strong on turf, routing, offline capture, and coverage, and they are expected to pass data to whatever system holds the customer or voter record.

Most poor purchases come from choosing the wrong category rather than the wrong product. A team whose real problem is that turf takes two days to redraw will not be helped by better automated sales tracking, however good the demonstration was.

The practical questions are few. What happens to a disposition entered with no signal? How long does it take a manager to redraw a boundary? When does the manager see the data? What does it cost when the team triples in size for the season? Those four questions narrow the options faster than any feature comparison chart.

The metrics that are hard to inflate

Doors knocked is the default measure and the worst one, because every easy way to raise it harms the program. Skipping the not-home record, walking the easy street again, and cutting conversations short all make the number go up.

Use these measures instead.

• Coverage. The share of the universe attempted at least once. Walking the same streets again cannot inflate it, because a second attempt at the same address does not increase it.
• Contact rate. Conversations divided by attempts. It shows problems with list quality and time of day. It stops being meaningful if not-homes go unrecorded, which makes it a useful check on data accuracy as well as on performance.
• Outcome per conversation. IDs, appointments, or signatures divided by conversations. It measures the script and training separately from the number of doors knocked.
• Attempts per canvasser hour, by turf type. Never combine turf types in this number, because comparing a rural canvasser to a suburban one on it drives out your rural canvassers.

Working out whether it paid

To judge the return on investment (ROI), cost per door is the wrong measure. Calculate cost per conversation and cost per outcome, and include recruitment and training in the cost. In a program with high turnover, those costs are larger than everything else, and they stay hidden if you only count hours walked.

For commercial teams, credit closed sales to the territory as well as to the rep. Reporting only by rep blames people for the quality of their territory and rewards whoever was given the good turf. It also gives you no way to tell a good rep on bad turf from a bad rep on good turf. Those two look the same on a ranking of results but need opposite responses.

For political programs, it is hard to know how many votes a canvass produced. The most useful substitute measure is coverage of the identified universe, rather than any attempt to prove that the canvass caused a result. Keeping turf inside precinct lines at least makes it possible to compare results afterwards.

Common questions

How big should a canvassing territory be?

Size by walking time, not door count. A territory is what one canvasser can attempt in one shift, commonly 60 to 120 doors on suburban streets for a three-hour shift and far fewer on rural routes. Equal record counts produce shifts of very different lengths.

Does route order actually matter?

Yes, and it is the cheapest available improvement. Ordering a walk list by street removes backtracking, which is worth several attempts an hour across every canvasser and every shift.

What should a field director look at during a shift?

Whether anyone has stalled, whether any turf is far off pace, and whether dispositions are being recorded at all. All three can be acted on only while the shift is still running.

How do you measure canvassing ROI?

Cost per conversation and cost per outcome, with recruitment and training included in the cost. For commercial teams, attribute revenue to territory as well as rep so the two can be told apart.

What is the most common canvassing operations mistake?

Managing to doors knocked. It is the easiest number to raise and the easiest to raise in damaging ways. Coverage and outcome per conversation are hard to inflate.

Where to go next

For a campaign, these decisions are covered with election deadlines, volunteers, and GOTV (getting out the vote) in mind in the guide to political canvassing and GOTV. For a sales team, the costs, earnings, and follow-up are covered in the complete guide to door-to-door sales.

Beacon is built for the work this guide describes. It offers territory management, route optimization, offline capture, and live coverage reporting, and a boundary can be redrawn in the middle of a shift.