Intro
Distributed organizing teams are built to scale quickly. To have more conversations with voters, at higher quality, faster than any single other piece of campaign infrastructure can be deployed.
The Biden-Harris campaign’s distributed organizing team was no different. Built over the course of just four months, the Team Biden-Harris HQ Distributed Organizing team organized 200,000 relational conversations, sent 150 million texts, made 88 million phone calls, and organized 11,000 supporter-led events. Put another way, the 28-person distributed organizing team, 1.5% of the campaign’s total organizing staff footprint, led 33% of the campaign’s total voter contact.
The concept of distributed organizing is not new to this cycle, but this was the first time distributed organizing has been used in a presidential general election campaign. And with the combined challenges of organizing during a pandemic and scaling our program at breakneck speed, the Biden-Harris campaign embraced the distributed model. Our distributed program built on and advanced programs that had been built over the past four years, including those programs built by staffers who would join the Biden-Harris Distributed general election team, programs like the Bernie 2016 and 2020 Distributed programs, the Warren 2020 Distributed program, and the Pete Relational program (itself built on programming created by Marshall Ganz for the Obama ‘08 campaign).
One strength the Biden-Harris Distributed Organizing team benefited from greatly was that our staff came from a number of 2020 primary campaigns. Our HQ Distributed team leadership came from the Warren, Bernie, Steyer, Pete, and Biden primary campaigns, which helped our team quickly adopt best practices from the many different teams who led this work throughout the cycle, and to be able to quickly collaborate on how different campaigns had approached distributed organizing challenges.
But for all of the buzzwords and changing technologies associated with “Distributed organizing,” distributed is not something distinct from organizing: distributed organizing is part of a long organizing tradition in America, from Huerta, to Alinsky, to Gompers, and countless more, that stretches back centuries.
Distributed organizing programs, as this post will explain, represent a radical vision of volunteer organizing, an evolution of caucus organizers and neighborhood team leaders for the internet age. A decentralized organization to train and empower hundreds of thousands of volunteer organizers to achieve goals otherwise impossible with only paid staff.
But first, to briefly define, what is distributed organizing? Distributed organizing teams are decentralized groups of individuals working together to achieve a defined organizing objective. Distributed teams have, as hallmark, three key calling cards.
Volunteer leadership. Distributed teams have, as general practice, large numbers of volunteer leaders, working in conjunction with a small number of paid staffers. This volunteer leadership structure is absolutely essential to building large voter contact programming with a small staff footprint, and is the number one aspect that defines a distributed organizing program. For practitioners, click here to read through the list of ways the Distributed team broke out volunteer leader roles.
Act now place agnosticism. Distributed teams take an act now, shiftless approach to volunteering. Volunteers can participate in a distributed program from wherever and whenever they’d like. No matter where you live, you can get involved with a distributed program. Further, where more traditional organizing programs tend to rely on a shift model, asking a volunteer to show up at a predetermined shift time (e.g. 2pm ET next Saturday), distributed programs allow volunteers to join and participate whenever they’d like, keeping volunteer opportunities available as close to 24/7 as possible.
Repeatable, automated, processes. Because distributed teams have a far lower staff to volunteer ratio, distributed teams generally rely on tools to help automate as many aspects of the distributed program as possible while building structured, repeatable processes so that one volunteer leader, or organizer, on duty can help provide direction for thousands of active volunteers. It is this tendency towards tool use to help solve problems that has led distributed programs to be associated with new organizing technology, like peer to peer text messaging, various Relational organizing apps, dialer tools, and more.
Put another way, the goal of the distributed model is to empower large numbers of volunteers to act in decentralized unison towards a common purpose, with processes and leadership to support the creativity members of the distributed team show as they work, frequently independently, to achieve the group’s goal. If a traditional state staffed organizing structure resembles the team behind the Encyclopedia Britannica, a distributed organizing team resembles a team like Wikipedia’s. Both encyclopedias have writers and editors, but while the Britannica team relies on a discrete set of staffers with highly centralized editorial control to dictate article assignments, the distributed program relies on a collective that grows as large as possible, and expressively encourages volunteers to contribute in as many ways possible that adhere to a set of community guidelines, using community volunteers to help create and enforce guidelines. In the same way volunteer writers and editors are the most important part of Wikipedia, volunteers and volunteer leaders are the most important part of a distributed organizing team.
Given the scale of a distributed program, volunteer leaders take on the role of de facto distributed community organizers. Nearly every aspect of the distributed program at scale should, for the program to be successful, be delegated to trained, accountable volunteers. As an analogy that may be helpful for those familiar with traditional field organizing: With volunteers taking on the role of an organizer, distributed staffers most directly take on roles equivalent to Regional Organizing Directors and Organizing Directors, to help build systems, provide training, and offer technical assistance to support volunteer leaders working to organize. And in the same way organizers on political campaigns are regularly promoted and given greater responsibilities, so too were our volunteer leaders. Testifying to the leadership pipeline our distributed team worked to build, and the depth of knowledge distributed volunteer leaders developed, six of the Biden-Harris Distributed team’s 28 staff members started as volunteer leaders on a distributed program.
Distributed programs have a primary impact on direct voter contact, but have a secondary advantage of activating and training thousands of volunteer organizers who continue to organize cycle after cycle. It is no coincidence that the Biden-Harris team drew in volunteer leaders who’d played a part helping build the distributed programs for the Bernie and Warren primary campaigns, just as staff who’d helped build those programs joined the Biden-Harris campaign themselves.
The distributed organizing program did not, of course, operate in a vacuum. Our distributed program operated alongside and in collaboration with a robust, staff focused, states organizing program, led by Biden Organizing Director Kurt Bagley, tasked with managing, overseeing, and leading teams made up of thousands of individual, paid organizers working in communities throughout our battleground states.
Where previous large-scale presidential campaigns generally relied on either a distributed or staffed organizing model to organize in key states, the Biden campaign ran the largest distributed program that’s ever been put together in Democratic politics, alongside one of the largest staffed organizing efforts that has been assembled. Both programs this cycle, and indeed, Democratic campaigns broadly, benefited from a level of volunteer interest higher than ever before. The National Distributed program alone recruited more than 160,000 volunteers, who completed hundreds of thousands of volunteer shifts, and served as organizers in their own right throughout our distributed community.
Because of this, the Biden-Harris campaign had the capacity to contact every voter in all 17 battleground states, as frequently as was necessary. The primary challenge was not how to recruit the most volunteers, though that certainly played a role. The core challenge was how best to empower volunteers to engage in the most impactful work. In our scenario, the distributed organizing program was set up, and is ideally set up, to supplement, not replace, on the ground, local organizing efforts. Working in conjunction with state-based organizing teams, Distributed organizing can serve as the large-scale, mass mobilization team, freeing state teams to focus entirely on community-based, relational work that is the strength of local organizing.
If ballots needed to be cured in Georgia, or Pennsylvanians needed to be reminded about changes in voter laws, for example, our distributed team could spring into action to make sure voters had the information they needed, freeing up capacity for local organizers to keep developing deep-rooted community networks.
While the Biden-Harris campaign Distributed team focused almost entirely on voter contact, distributed programs can support any large, repeatable task campaigns need help with. Distributed teams can, for example, help translate voter contact scripts, produce debate transcripts, cut door-knocking turf for Get out the Vote programs (when not in the middle of a global pandemic), or any other repeatable, trainable process requiring a significant investment of time and labor. The only limitation is in a campaign’s creativity.
With this as a prelude, what did the Biden-Harris National Distributed program do and how did we do it?
The Biden-Harris National Distributed Team
The Biden HQ Distributed program was made up of four main verticals, with additional support from Deputy Distributed Director Spencer Neiman and our distributed embeds — Distributed Analytics Director Nina Wornhoff, Distributed Analytics Analyst Shareq Rashid, and Deputy National Training Director Lucia Nunez. Those verticals included:
The Call Team, led by Claire Wyatt, staffed by Erin Kwitny, Kevin Stephen, Melissa Goldberg, Graham Derfner, Bridget Wu, and Isabel García-Ajofrín.
The Calls team organized our distributed call volunteers via Slack, our online organizing hub and “virtual field office.” Starting in July 2020, Calls Team recruited, trained, and organized the largest distributed calls programs that has been put together in Democratic politics. The team ran a bilingual, English and Spanish, calls program (through our Call Crew and Spanish language calls program, Llamando Contigo) that made more than 88 million calls in just four months. Previously, the largest distributed calls program was organized by the Bernie 2016 campaign, which made 75 million calls in a little over a year (and of which several of us are proud alumni).
Claire and her team ran a bilingual, English and Spanish, calls program. As a part of our Call Crew, Isabel García-Ajofrín designed and ran ‘Llamando Contigo,” the campaign’s Spanish language calls program. Isabel cultivated a community of callers who made millions of calls to Spanish-speaking voters in key battleground states including Florida, Pennsylvania, Arizona, and Nevada.
During COVID, our calls program was one of our most effective ways to hold conversations with hard to reach voters across the battleground states. Because of this, Claire and her team needed to build an intentionally state-specific program. As the program grew, her team developed state-specific scripts for all 17 battleground states, including unique scripts for each phase of the campaign. This modularity of scripting required a tremendous amount of work on behalf of Claire and her team, but was worth it to make sure volunteers had the most impactful conversations possible. Additionally this helped build trust among our volunteer organizers — we were respecting their time by making the most impactful ask of them, and helping volunteers serve as informed community organizers, mobilizing voters with the most up-to-date voting information.
Pictured above, a staff announcement Slack post announcing calls into Georgia
Claire and her team embraced the distributed ethos of using the best tools available to have the largest possible impact in the shortest period of time. Calls team used a dialing tool, ThruTalk, which saved volunteer time by allowing volunteers to speak with voters without dialing numbers themselves. While this increased the training demand on her team, the results spoke for themselves. Where a volunteer organizer on the Obama ‘08 or ‘12 campaign might be able to make 45 to 60 calls an hour, a volunteer on Claire’s team could make upwards of 300 calls an hour.
Calls Team helped build culture through their caller training program. Would be callers were highly encouraged to attend one of the Call Team’s live trainings before they started making calls with the team. The Calls Team and their team of volunteer training facilitators offered trainings 3-7 times per day, and made training the major entry point of the program. During this training, facilitators not only taught new callers how to use the campaign's dialer tool and how to participate in the Distributed Slack caller community, but also shared the program’s theory of change, invited volunteers to connect Joe Biden’s values with their own life stories, and taught volunteers how to have effective conversations by “relating, not debating” with voters. Call team trainings would, for example, emphasize the “field margin” the 2016 Presidential election came down to, so as to drive home the importance behind the voter contact work the team was doing. This training significantly helped build volunteers' buy-in and connection to the program and volunteer team.
Claire and her team placed an incredible emphasis on building a vibrant virtual volunteer community through their Distributed Slack channels, which surged to over 103,000 volunteer callers by Election Day. The success of the calls program spoke to the importance of building an active, engaged Distributed Slack community. Claire and the Call Crew asked volunteers to join a team, not just perform a task, and invested deeply in cultivating an atmosphere in the #call-crew Slack channel that was highly energetic, effusive, supportive, fun, bubbly, and constantly buzzing. Those cultural norms helped volunteers stay invested and return shift after shift, replicating some of the magic of organizing in person together.
Claire and her team created Slack norms to keep volunteers active on Slack every time they called with the team, deepening their connection with each other and the program. Call Crew volunteers constantly posted cheerful “sign on” messages to say hello when they arrived, asked questions as needed throughout their calling, posted specific animated emojis when they had a successful call, cheered on other callers, and “signed out” by sharing how their calls went, any interesting conversations they had, any difficulties they encountered, and when they’d be back. These norms helped connect volunteers to leaders and to one another, and kept volunteer callers coming back.
Slack served the dual purpose of culture building and responding to questions. The 300 person team of volunteer Call Crew Slack moderators who helped lead the #call-crew Slack channel responded to volunteer questions near-instantaneously during the 14 hours of volunteer calling every day, at peak, answering a new volunteer question posed in Slack every 3 seconds.
Because of the culture that Claire and her team built, volunteer callers spent an average of 45 minutes using the campaign’s dialer each time they logged on to make calls, up to 50% longer than on comparable programs.
The Relational Team, led by Roohi Rustum, staffed by Trevor Fifer, Kai Mateo, Victor Rubio Rivera, Kathryn Short, and Erin Stevanus.
Where many parts of organizing programs ask supporters to make calls or knock on the doors of strangers, relational programs work by asking supporters to organize those they already know to take action with the campaign. Through a relational program, volunteers would be asked to organize, for example, their high school teacher, an Aunt, a former coworker, e.g., living in key battleground states. This is often the campaign’s best way to reach hard-to-reach supporters, first-time voters, or other folks not traditionally contacted by campaigns. Research shows conversations with our friends and family about politics is one of the best methods we have to increase support for our preferred candidates and policies.
But having conversations with our friends and family about politics is difficult for many of us. There is a reason, after all, that Thanksgiving political talk gets such a bad rap, and indeed, relational programs that have asked volunteers to take this action by themselves have historically struggled to make a significant impact. It’s hard to have conversations with the folks you know about politics, which is why our relational team put training and support in the forefront of our relational work.
Roohi’s relational program shone by asking volunteers to take action together with other volunteers, rather than taking action alone. Roohi and her team developed deep volunteer leadership, asking volunteers to organize their own communities through values-driven conversations. Staff hosted friendbanks (like a phonebank, but where you reach out to your friends), Strategy Sessions, and Vote Joe (our relational organizing tool) trainings to recruit the first wave of volunteer leaders, and then, trained volunteers to host their own relational shifts. Roohi and her program built a, “People first, app second,” relational program, focusing on organizing training before training on the app, since starting with organizing training first makes shifts accessible to those who struggle with apps and lays the groundwork to introduce the app later as volunteers become excited to track the work that they completed. This helped increase app adoption, which is frequently a pain point for relational organizing programs.
Roohi and her team then reinvented relational metrics, which have often relied on “messages sent” or “user contacts upload to an app”, to instead track volunteer leaders, voter contact shifts, and finally, reports filed on the number of conversations our volunteers had and what they learned through their conversation — how is your contact voting? Who are they voting for? Tracking reports on conversations vs. just “messages sent” was critical to ensure conversations were actually happening — and empowered volunteers to reach out to their people however they felt comfortable. Volunteers didn’t need to use Vote Joe to have their conversations. They could do so via phone, social media, or text message, and simply use the Vote Joe app to report back to the campaign.
The introduction of “Green Star” voters was one of the Biden-Harris Relational program’s biggest innovations. Through our relational app, volunteers could sync their phone contacts to quickly identify which of their contacts were on the campaign’s priority outreach lists, marked in the app with a Green Star. We uploaded two sets of priority contact lists to supporters. In September, we showed supporters their contacts in battleground states (by matching supporter’s contact area codes with the national voter file), to encourage supporters to try relational organizing and to highlight that our supporters likely knew more people in key, battleground states than they realized. Then in October, we rolled out the new “Green Star” feature to match and highlight supporter’s contacts with voters who had requested, but not yet returned, their vote-by-mail ballot or who were in the campaign’s GOTV universe (i.e. the supporters we most prioritized turning out).
The volunteer would then be provided with key voting information to share with their connection. For example, a volunteer might see their sibling in Wisconsin hadn’t returned their mail ballot, or a friend in Pennsylvania hadn’t voted early in person yet. This was an incredibly effective tactic for the campaign: In the final stretch, 84% of our relational conversations were held by supporters in non-battleground states to supporters they knew in battleground states.
The Biden-Harris Green Star program dramatically increased use of the Vote Joe app and the number of relational conversations our volunteers had. By building volunteer skills, having tough conversations with people they knew first, and demonstrating the app’s value as a tool to show volunteers their highest priority contacts second, after we introduced the Green Star voters program, volunteer report backs on their relational conversations went up nearly 300%.
A graph tracking volunteer reports filed on relational conversations. The 300% spike on 10/23 corresponds to our introduction of “Green Star” voters.
In the final weeks of the election, Roohi’s team also partnered with the Biden-Harris coalition team’s Young American Engagement Director, Hannah Bristol, to run a paid relational Fellowship to encourage supporters to reach out to their networks on college campuses and other harder-to-reach communities. A key difference from prior paid relational work was that we compensated Fellows with a stipend, rather than paying hourly wages.
Compensating supporters for “texts sent” or “contacts made” is difficult with relational voter contact: if you text your Mom during an hourly shift Monday and she responds Thursday, she’s responded outside of the shift. There’s no structural incentive to follow-up to have that conversation about voter registration, candidate support IDs, or a voting plan, let alone incentive to track that conversation. To solve this problem, we created a DNC Fellows program, providing Fellows with a stipend during the final weeks of the campaign.
We provided the same training resources, development, and support that volunteer leaders, organizers, and digital organizing directors received. Fellows then ran their own friendbanks to organize their communities in the final weeks of the campaign. On average, our Fellows reported back an average of 32 conversations per Fellow, ten times more conversations than a supporter participating in a friendbank, and six times the number of conversations as a traditional phonebank volunteer.
Roohi’s program contained countless innovations, detailed through her team’s “Relational Playbook” and resource directory. For practitioners, request resources here.
The Text Team, led by Clarice Criss, staffed by Brianna Kirkland, Jack Arnheiter, Max Kennedy, Carolina Malagon, Joe White, and Victoria Pinilla.
Clarice’s Text Team was the engine that powered our National Distributed organizing program. Clarice’s team and volunteers sent 150 million peer-to-peer text messages in both English and Spanish through our Text Team and Spanish language text program, Text Mi Gente. While Text Team was tasked with direct voter contact, particularly in large battleground states with smaller staff footprints, most battleground state distributed organizing teams had the necessary capacity to meet their texting needs through in-state volunteers. Because of this, the tens of thousands of people who made up our National Text Team served to primarily recruit and mobilize voters and volunteers in the 33 non-battleground states, with a focus to recruit and train on key methods of voter outreach such as making calls and relational organizing.
Distributed, peer-to-peer texting has evolved significantly since 2015, when Bernie Sanders’ first presidential campaign started using the tactic. During the 2016 primary, the Sanders campaign sent 8 million text messages. During the 2020 general election, the Biden-Harris campaign sent more than 330 million. While the Bernie campaign was one of a handful of campaigns who embraced P2P texting in 2016, by 2020, texting has become a widely used tactic.
With increased P2P activity across the campaign space, our Distributed Text Team operated in a noisy texting environment. The challenge our team faced was not how to ensure that every voter got a text, but how to make sure voters and volunteers got the right text at the right moment. Our team sent texts asking unregistered voters to register to vote, provided updates on changes to voting laws, reminded voters to return their vote by mail ballot, and of course, helped voters find their polling place. Clarice’s team also recruited volunteers to make phone calls, host events, organize their friends and family, staff voter protection hotlines, and more.
At the peak of our program, the Distributed Text Team launched a new, hundred-thousand person texting campaign every sixty seconds. To do this, our team worked with the Biden-Harris tech team to design a number of new tech tools, including an automated campaign creator (the CLR), to automatically build peer to peer texting campaigns from a Google spreadsheet, “magic links,” to automatically add trained texters to new texting campaigns, and a tool to automatically reassign text messages that came in from previous volunteer shifts that still needed a reply.
Pictured above, the page volunteer texters would see to fill out when they clicked on a magic link
Our team worked with the Biden-Harris Tech Team to create tools to streamline the texter onboarding process. ROB Bot, a Slack bot that walked volunteers through our onboarding process automatically, where previously, volunteer greeters would onboard volunteers on a one by one basis. ROB Bot helped make sure that texters had completed our volunteer training, texter quiz, and created their texter account before they had access to the Slack channel where we shared texting magic links to help keep volunteer texter quality high. ROB Bot could even flag volunteer organizers through our #mods-rob-help channel when a volunteer texter was having trouble getting started.
Pictured above, ROB Bot hard at work onboarding new texters to our Text Team
Our team recruited “front end” volunteer leaders to answer texter questions via Slack and help onboard new texters who couldn’t make it through our automated onboarding process, and backend volunteer leaders, who worked to assign texts to volunteers, reassigned texts that hadn’t yet received a reply so the messages could be sent out, read through texter responses and gave feedback as necessary to help make sure our volunteers improved as texters, reassigned text messages to special, language volunteers help reply to text messages in more than 20 different languages, and more. Thanks to the incredible work of the Distributed Text Team, anytime anyone responded to one of the 150 million individual text conversations our team sent, they received a reply back in, on average, less than 10 minutes.
Above, a heat map representing the messages Text Team sent to every state in the US
The Events Team, led by Catherine Vegis, staffed by Serena Bian, Reagan Hunter, Angela Elder, Charlene Wang, and Hannah Peyton.
Catherine Vegis and her team led a voter contact-focused, volunteer-led events program. They coached, supported, and trained volunteers to host their own phonebanks, friendbanks, watch parties, and more. The purpose of this team was to help drive voter contact, both in terms of the community members our hosts mobilized, and the voters contacted at the events themselves. While debate watch parties and community events have their value, Catherine and her team emphasized the importance of voter contact as the team’s overriding purpose, to make sure our volunteers were asked to do the single most impactful thing possible.
As a secondary purpose, Catherine’s team also led large, all volunteer events. Her team led volunteer calls each night during GOTV (the final four days of the campaign), large volunteer thank you calls, staff-led Phonebanks, all designed to ensure our volunteers had high-quality events to attend every day of the campaign.
Volunteer hosts take on a large organizing responsibility, and Catherine and her team worked to make sure our hosts had the training, guides, and support from staff to encourage volunteers to take that next step and hold their own event. Catherine and her team built out resources, like a Rapid Support Center, a volunteer leader-staffed online Zoom room where volunteers could join if they had questions at any point, and “How to Host an Event” trainings and guides. Because new volunteer hosts can often struggle to recruit enough attendees to make their events a success, Catherine and her team made sure to support our volunteer hosts with event recruitment tactics, like large-scale peer-to-peer text email recruitment support, to make sure first-time event hosts had successful events to encourage these volunteers to host again.
But, volunteer hosted events were hosted on the Biden campaign’s events platform tool, Mobilize America, meaning they were fully public. Because of this, Catherine and her team worked to make sure that every volunteer submitted event was reviewed by a volunteer leader. Catherine and her team ran a, “Mobilize Moderator,” program, to make sure that volunteers reviewed each and every one of the 11,000 volunteer hosted events in non-Battleground states.
Above: The Distributed Organizing Staff Org Chart
Conclusion
The distributed team could not have achieved the scale it did, making 88 million calls, organizing 200,000 relational conversations, sending 150 million texts, and supporting 11,000 supporter led events, organizing 33% of the campaign’s voter contact, with just 1.5% of the organizing staffers, without training and activating tens of thousands of volunteer organizers to mobilize 160,000 organizing volunteers.
Overall, the distributed team held direct voter contact as the primary goal of our team, treating our voter contact as a decentralized force that the campaign could rapidly deploy to help the states where organizing support was most needed. As a series of principles, distributed organizing programs organize and train teams of volunteer leadership, empower volunteers to act now from wherever they may be, and build out new technical tools and solutions to engage in big, structural organizing.
While the line between “Distributed” and organizing is ultimately artificial, the impact of organizing programs that radically trust and empower volunteer leadership is anything but. Distributed programs should not be looked on as something distinct from the long tradition of organizing. The practice, and the vision, of hundreds of thousands of volunteers and organizers making up a distributed program, is just one part of a long organizing tradition.
When an organizer empowers a volunteer to help organize in their communities with a defined group, organizing is taking place, whether that conversation happens on a phone call, in an office, in a Zoom room, or in a Volunteer Slack.
The Biden-Harris National Distributed Organizing team helped the campaign reach record-breaking levels of voter contact in a very short period of time, in a scaling period not possible without developing tens of thousands of volunteer organizers to support the effort. This more than any other aspect is the legacy of the distributed program, the training and support to help first time volunteer organizers learn the aspects of organizing that mattered most, and the creation of 160,000+ volunteers, leaders, and organizers ready to organize in their own neighborhoods for cycles to come.
To every single volunteer who helped build this team, and this program; who joined this effort to advance progress; to those who started learning to organize for the first time, and to those who have been organizing for years; to every single person who made a phone call, sent a text, hosted or attended an event, who talked to their friends and family about the importance of this moment. Thank you. This is your victory.
Nathan — thank you so much for this incredible documentation of what you and so many others created. I have the privilege to serve as Movement Voter Project's Director of Donor Organizing, and your work is an inspiration to me as I begin to build out our program!
Alinsky did something entirely different: A constant community presence, not a drop in 4 months before the desired event. Talk of "state-based" programs emphasizes the point -- community organizing is what is important. The DNC needs to find a way to fund year-round community-based efforts. Prediction: By 2024, 90% of Democrats will have figured out how to text back "Stop."