Best Practices for Canvassing—Improving Contact and Turnout

The Digital Ground Game Newsletter

With the midterms fast approaching and the majority of consequential Senate races having been covered, Digital Ground Game aims to make some adjustments to our recurring newsletter. Our goal is shorter but more frequent publications; ideally, this will provide a more digestible product that's easier to keep up with. As our canvassing operations ramp up, the scope of the reports will expand to include best practices and tutorials on things like door knocking and advocacy more generally. Not only should this benefit our own canvassers, but its broader application also circumvents the local nature of election coverage, making it more palatable for a nationwide audience. For those who enjoyed my election coverage, the current plan is to alternate between covering new topics and providing updates on the remaining important Senate races, along with our path to the Senate overall. Now that we've outlined the changes to the newsletter and the broader focus, let’s shift to effective strategies for canvassers to increase their contact rate and increase voter turnout. In each section, I’ll cover strategies from some of the larger door-knocking organizations nationwide and pick apart the reasoning behind them, but first I’ll lay out some terms to paint a picture of the goals these strategies aim to achieve.

Increasing Contact Rate

The contact rate of an operation refers to how often residents will come to the door and talk to canvassers. Contact rate strategy is a pivotal element of ground game and in some cases, can be the determining factor in whether an operation succeeds or fails. What encompasses such a strategy can be quite broad, but it might include things as simple as ensuring door knockers avoid meal times, or as complex as utilizing canvassing apps to identify neighborhoods with high swing voter density ^[1]. While some tactics are locally tailored, many can be utilized across the country with some minor tweaks.

Increasing Turnout Rates

Turnout refers to whether the residents contacted ultimately decided to vote or to stay home. Similar to contact rates, turnout strategies can vary greatly depending on the locations or demographics canvassers engage with. In simpler circumstances, increasing voter turnout looks like registering people to vote or getting existing voters excited about upcoming elections. In more difficult dynamics, it involves convincing oppositional voters to come over to your side by humanizing yourself through recurring face-to-face conversations or friendly debate. Whether debate is allowed varies from group to group, so be sure to check with organizing directors before attempting this.

Swing Left

Given their strategic similarity to Digital Ground Game, both focusing heavily on toss-up races, opening with Swing Left ^[2] feels appropriate. Swing Left’s ground game strategy is a program called Ground Truth ^[3][4], which it describes as “[a strategy] that starts earlier, listens deeply, and expands the universe of voters [they] engage.” The program was launched in response to the Democratic Party’s failure to reach low-propensity but persuadable voters in 2024; as a result, it differs from traditional canvassing strategies in a handful of significant ways. Likely the simplest change is Ground Truth’s initiative to “knock every door”; while conventional wisdom is to use voter data to focus on bolstering turnout in existing Democratic voters, Swing Left aims to expand the party’s voter base by making contact with residents who might otherwise be passed over due to party support scoring metrics ^[5]. In line with the “starts earlier” section of the tagline, Ground Truth’s volunteers begin making contact with voters over a year ahead of elections. This change supports the goal of deeper conversations by allowing time for a more reliable framework of repeat contact, strengthening rapport with voters. The key, and most complex, change is the project’s shift away from the quicker, survey-based contacts traditionally seen in door knocking in favor of longer-form, open-ended contacts. This would be difficult to replicate without a longer timescale, but some elements could likely be adapted to our benefit without requiring major structural changes. Canvassers are instructed to take detailed notes on the exchange as well as record a brief reflective voice memo after each interaction. An AI model processes all of this and generates extensive datasets used by Ground Truth to calibrate the focus of future voter contacts. Additionally, the insights on constituent interests gathered this way can currently be sold to candidates directly based on the 2024 FEC advisory opinion on paid canvassing coordination ^[6].

Turning Point Action

In order to form a well-rounded perspective on popular strategies here, we’ll jump to the other side of the ticket and look at Turning Point Action ^[7], the wing of TPUSA that handles its grassroots organizing efforts. Turning Point Action’s equivalent to Ground Truth is a program called Chase the Vote ^[8]. Chase the Vote has a slightly longer track record than Ground Truth, having been launched ahead of the 2024 elections, but interestingly shares some strategic themes. Most notable is the critique of “transactional” survey-based voter contact, which it identifies as the previous M.O. of the GOP’s ground game operations. Chase the Vote’s exact approach isn't public, but its rhetoric surrounding this aspect of the program comes off similarly to Ground Truth’s open-ended dialogue. While Turning Point Action doesn't start its operations as far ahead of elections as Swing Left, its answer to the issue similarly favors open-ended contacts with a focus on building connections with voters. This is largely where significant strategic overlap ends; from here, Chase the Vote mirrors traditional organizing efforts more closely, making some of its tactics potentially easier to integrate on a tighter timeframe. While it still targets low-propensity voters, Chase the Vote focuses on residents who are already ideologically Republican or pro-Trump, thereby reducing the number of wasted contacts. The organization’s “ballot chasing” strategy seems to make it stand out from contemporary activist groups. Using its proprietary app, canvassers keep track of the contact info of voters they’ve connected with to remind them of outstanding mail-in ballots or upcoming elections ^[9]. This not only strengthens connections with voters, but allows Turning Point Action to keep internal contact engagement records akin to party support scoring metrics used by similar groups.

Strategy Synthesis

In this section, I’ll cover strategies seeing wider adoption, how our Digital Ground Game teams use these tactics, and how each strategy seems to impact contact and turnout rates. The first thing a canvassing group needs to do is determine what houses to visit; generally, this is done either using a proprietary app or by purchasing/requesting voter data from the state. Our team uses voter data to focus on neighborhoods with a higher density of Democratic Party voters ^[5], allowing us to focus on improving the turnout of existing voters, which is generally considered more time-efficient ^[10]. Collecting the contact information of the voters canvassers encounter allows for reminders to be sent out regarding things like upcoming elections and the distribution of mail-in ballots. A 2025 Center for Campaign Innovation study found that sending just one follow-up text after a canvass visit increased turnout by almost two points, though the effect was concentrated almost entirely among women, who saw a roughly four-point boost ^[11]. Digital Ground Game already has access to the necessary contact information and is weighing how best to deploy such a program. Larger canvassing organizations seem to generally be pulling away from more stilted, survey-based voter contact formats, citing evidence that longer, more substantive conversations can more durably shift voter attitudes ^[12]. The specifics of our conversational scripts are determined by the leadership at each of our canvassing houses, but implementing more free-flowing guidelines is something our teams are actively working on. Ideally, by leaving voters with a more favorable impression, we can improve turnout. Some groups have begun exchanging data collected by canvassers directly with candidates based on the aforementioned 2024 FEC advisory opinion ^[6]. Given the relative newness of the opinion, only a handful of groups do this, and it's still unclear what overall effect such a strategy will have on campaigns. As such, Digital Ground Game has opted not to implement this without further evidence of its efficacy. Last is a strategy that smaller groups will generally be priced out of: the longer operational timeframe seen in Swing Left’s Ground Truth program, employed to more solidly establish the organization in communities. While this tactic seems promising so far, the cost of adding months to a canvassing project is too prohibitive for smaller organizations like Digital Ground Game to attempt without more data on the program’s effects. That said, leadership has expressed interest in extending operation times in the future, so keep an eye out for updates on the Digital Ground Game website.

Author's Note

Because this newsletter doesn't have my usual extra reporting section at the end, I wanted to add a less formal addition here for readers who have kept up with my election coverage. For those who haven't heard about the Democratic primary for the Wisconsin Governor seat, DSA candidate Francesca Hong lost to the more moderate David Crowley by less than 1%. This was initially a shock (including to me, embarrassingly) because reporting ahead of the primary showed Hong maintaining a significant polling advantage in the race. The reality was that, while Hong had a lead over other named candidates in the polling, a significant percentage of respondents were undecided, and the remainder was split between a handful of more standard moderate Democrats ^[13]. Because polling aggregate sites generally don't list “undecided” as the leading candidate, people who don't open the polling breakdowns would just see a list of polls where Hong has a significant lead. I was kicking myself when I realized this because I wrote about a near-identical situation happening in the Alaska gubernatorial race back in my June report, where an abundance of Republicans split the poll, making the only viable Democrat appear to have a significant lead in a fairly even race ^[14]. This is just further evidence that polling reporting needs to improve so confusion like this can be avoided in the future.