Track Your Canvass Week by Week, or the Numbers Tell You Nothing
A cumulative total always goes up. That is precisely why it tells you nothing.
A campaign that has recorded 12,000 contacts feels ahead of one that has recorded 9,000. Neither number says whether the work is speeding up, slowing down, or stopped three weeks ago while the total sat there looking healthy.
A total is not a rate
The total answers how much has been done. The rate answers whether it is still being done, and only the second one can be managed.
Suppose a campaign records 400 contacts in its first week, 380 in the second, 390 in the third, then 250, then 180. The total is still climbing. It has never once gone down. And the campaign is failing, visibly, from week four onward, to anyone who is looking at the weekly figure rather than the cumulative one.
Those numbers are invented for the illustration. The pattern is not.
What to record every week, in the same shape
Five figures, per ward, every week, without exception.
Contacts attempted. Doors knocked, calls made.
Contacts made. People actually spoken to. The gap between attempted and made is a fact about your area, and it is stable enough to plan around.
Identified supporters. Of those spoken to, how many said yes.
Refusals and no-answers. Recorded separately. A no-answer is a return visit. A refusal is not, and confusing the two wastes weeks.
Active canvassers. How many people actually went out. This is the divisor for everything above.
The same five, in the same order, every week. Changing what you count is the fastest way to lose the ability to compare, and comparison is the entire point.
Related: what a ward coordinator is measured on.
Reading the trend without over-reading it
Three patterns, and what each usually means.
A steady rate with a steady supporter share. The operation is working. Leave it alone and extend it to another ward.
A falling contact rate with a steady supporter share. This is a capacity problem, not a message problem. Either canvassers stopped going out or the easy streets are finished. Look at active canvassers before you touch anything else.
A steady contact rate with a falling supporter share. This is the important one. Either you have moved into less friendly territory, which is expected and fine, or something changed in what people are saying to you. Ask the canvassers before you conclude anything. They usually know, and nobody asks them.
Compare like with like, or the trend is fiction
Three traps that make weekly figures lie.
Different areas. A week spent in your stronghold and a week spent in a contested ward are not comparable, and a drop in supporter share between them means nothing at all. Record the ward alongside the number so the comparison is possible.
Different weeks. A week containing a public holiday, a funeral that stopped the ward, or heavy rain is not a normal week. Note it next to the figure. One line of context saves an argument in a month's time.
Different definitions. If one coordinator counts a household as a contact and another counts each adult in it, their numbers cannot be added. Write the definitions down once and hand them out.
Related: going door to door at county scale.
Why this cannot be done from memory
A campaign of any size has several wards, several coordinators and several weeks. That is a grid, and grids are not held in heads.
In a ward you can keep it on paper. Across a constituency the paper stops working around the point where you need last month's figures during a Saturday meeting. Across a county it was never going to work, which is why county campaigns that run on recall discover in July 2027 that two sub-counties stopped canvassing in April and nobody noticed.
Do this at your next Monday meeting
Write the five figures for last week on one sheet, ward by ward, next to the same five figures for the week before.
Then ask about any ward where the contact rate fell by more than a quarter. Do not assume you know why. Ask.
Related: the five numbers to read every Monday.
Frequently Asked Questions
What should a campaign measure when canvassing?
Contacts attempted, contacts made, identified supporters, refusals and no-answers kept separate, and the number of canvassers who actually went out. Record all five per ward per week in the same shape, because the value is in comparing this week with last week rather than in any single figure.
Why is a cumulative contact total misleading?
Because it can only rise. A campaign whose canvassing collapsed a month ago still shows a large and growing total, so the collapse is invisible on the figure everyone quotes. The weekly rate shows a slowdown immediately, which is the only version of the number a manager can act on.
What does a falling supporter share mean?
Usually one of two things. Either canvassers have moved into less friendly territory, which is expected and needs no response, or something has changed in what voters are saying. The way to tell them apart is to check which wards the week's work happened in and then ask the canvassers directly.