Living with VANxiety: The Present and Future of Progressive Movement Tech

How did the progressives come to be nervously dependent on core technology owned by an overseas private equity fund? How bad is this situation? And what can be done going forward? A special report.

[Note to readers: This week I’m taking a deep dive into the state of the technology that Democrats and progressives use to campaign and organize. If you are someone who subscribed in the last six months because of my coverage of the politics of Israel/Palestine here in America, feel free to take a breather. Or, enjoy the swim!]

“What does it mean to be infrastructure? We’re not the car driving down the highway, we’re the highway.” —Amanda Coulombe, former President of EveryAction/NGP VAN, speaking on The Great Battlefield podcast, April 2022

“My worst-case-scenario is a world where NGP VAN goes up for sale in ‘25 after they've had a very profitable year and it is purchased by a Elon Muskesque figure.” —Julia Barnes, CEO, The Movement Cooperative, November 2023

Every day of the week, hundreds of thousands of people get an email, a text message, a phone call, or a knock on the door from a progressive community organization or Democratic political campaign that found its way to them via a powerful suite of software tools. Whatever the entity doing the outreach, be it a nonpartisan voter registration outfit or an issue advocacy organization or a partisan campaign, and whether that entity is asking for a vote, or a donation, or a signature on a petition, they are almost certainly using one or more of these software platforms: NGP VAN, EveryAction, ActionKit or Mobilize.

These are some of the most important pieces of digital infrastructure powering the entire ecosystem of Democratic and small-d democratic political and civic engagement on the center-left in America—indeed for roughly 16 years NGP VAN was the de facto monopoly provider of campaign tech to Democrats. NGP VAN and EveryAction are critical for everything from fundraising to voter targeting to canvassing. ActionKit enables organizations with huge email lists to not only run massive online campaigns, but to segment their lists by Congressional district, something no other mass email tool does. And Mobilize is the main tool that campaigns, advocacy organizations and community groups uses to steer volunteers for phone-banking, text-banking, canvassing and other organizing events. With more than four million active users in 2020, there is nothing else like it.

Unfortunately, for the last two-plus years, many top leaders and strategists in this ecosystem have been quietly losing sleep over one worrisome fact: in August 2021, EveryAction, the tech conglomerate which is the source of this powerful suite of software tools and data, was sold to Apax Partners, an overseas private equity fund. Apax in turn merged EveryAction with two other companies (Social Solutions and CyberGrants) into a larger conglomeration of software services under a new company name, Bonterra, that is mostly focused on the giant nonprofit sector. And with interest rates rising and the economy slowing, Bonterra started shedding engineers and cutting back on support for these key products. In 2023 alone, it made across-the-board cuts of almost one-third of its workforce.

Some organizations that have built their programs around these tools are already in a panic about the changes underway, and there’s a general expectation that support for key tools like NGP VAN and ActionKit is going to inevitably decline as usage rises with the 2024 political calendar. Most Democratic and progressive technology experts believe Bonterra will do its best to keep these tools afloat and functioning through the end of the election. But after that, all bets are off. “In 2025, will this all fall off a cliff?” one top progressive tech leader asked rhetorically. “Definitely,” was their answer. Or as another put it, people are feeling “VANxiety.”

Chelsea Peterson Thompson, the general manager of NGP VAN, recently spoke with me on the record about these concerns and pushes back strongly against them. Asked about the staff cuts made by Bonterra that set so many alarm bells ringing, she says, “Last year, we, like many others, had to make the tough but strategic decision to reorganize our workforce to make sure our political customers have the exact resources they need in a critical election cycle.” She says Bonterra is committed to supporting NGP VAN into 2025 and beyond. “It’s fairly common to see anxiety in our ecosystem, particularly in an election year,” she told me. “We don’t expect election-related anxiety to go away. That being said, NGP VAN has always operated in order to prove ourselves every election cycle and will continue to do so.” She added, “Come fall, I certainly hope that our clients see the outcomes they expect and feel more comfortable with us.”

How did we get here? How did the progressive movement come to be nervously dependent on core technology owned by an overseas private equity fund? How bad is this situation? And what can be done going forward to change things for the better? Experienced technologists and campaigners across the progressive ecosystem have been discussing these questions in private for some time now. For the last six months, I’ve been interviewing a wide range of participants from across the field. Today, I am publishing an in-depth report, “Living with VANxiety: The Present and Future of Progressive Movement Tech,” as my contribution to surfacing that conversation into public view. It offers three things: a) a history of the evolution of NGP VAN from bootstrapped start-up to private tech conglomerate; b) a discussion of the severity of the challenges currently facing progressive organizations dependent on Bonterra’s tools; c) an exploration of the options going forward.

The full report, with an accompanying appendix describing the key tools at stake, is a certified “long read” at nearly 11,000 words. You can read it in full on my website. What follows is a compressed version.

Part One of the report (which I’m mostly skipping over here) charts the whole trajectory of NGP VAN from two scrappy startups that were pioneers in using tech to support political campaigns to becoming Every Action, a 500-person company that claimed annual revenues of “well over $100 million.”[1] That’s when Apax Partners gobbled it up for an undisclosed sum.[2] According to Bloomberg, the combined deal with CyberGrants and Social Solutions cost $2 billion, “including debt.”[3] Bloomberg also reported that the combined new company “will have a network of 650,000 non-profits, half the Fortune 500, and over 38 million donors and volunteers.” One thing it would not have was Stu Trevelyan, the longtime visionary CEO of NGP VAN who had steered it towards this massive roll-up; he quietly stepped down at the end of November 2021.[4] Over the following months, many of the other Democratic campaign professionals who had also worked alongside Trevelyan prior to the Apax deal would also depart for other jobs. And it’s this transformation, from a centerpiece of the Democratic campaign world to subsidiary of a foreign conglomerate, that is at the heart of today’s VANxiety. [To read more about the full blow-by-blow evolution of NGP VAN and how it came to be taken over by outside investors, as well as a discussion of whether the DNC missed an opportunity to keep the company from falling into private hands, go here.]

How Serious A Problem?

Part Two of the report explores the current situation. Knowledgeable observers do not agree about whether it is a five-alarm fire or something more manageable. Here are my main takeaways:

It is highly unlikely that Apax/Bonterra would sell NGP VAN or any of its related tools to an adversary of Democrats or the progressive sector because that would damage its desired reputation as a benevolent provider of tools for the entire nonprofit sector. The DNC says it is confident that the legal protections in its current contract with NGP VAN are sufficient to protect its interests and that it has taken the pro-active steps needed to reduce any risks or dependencies on outside vendors. That said, non-profit and advocacy organizations may be in a more vulnerable position.

That’s because Bonterra could decide to maintain the core NGP VAN product suite at least through the 2024 cycle and possibly longer, but allow other critical products, most notably ActionKit, to fall by the wayside. This scenario appeared to be underway last fall when half of the small ActionKit engineering team was let go, but after a public pressure campaign, two senior engineers were rehired. Chelsea Peterson Thompson says NGP VAN is “growing ActionKit” and “actively investing in it,” not only recalling employees but also opening additional roles and “making it a truly cross-functional product and integrating it across NGP VAN.” Still, organizations like MoveOn, which are heavily dependent on ActionKit, having built many customized operations centered on its platform, are taking no chances and being forced to invest heavily in finding or building alternatives.

Third, Bonterra could continue to maintain these core tools, but innovation, maintenance, and customer support will all decline. Thompson insists that “we have a robust roadmap leading into the election” and says new features will continue to be released across all their products “as we get closer to Election Day.” She also adds that they are steadily adding dedicated staff to the NGP VAN team. But of the nearly 30 experts I spoke to, the general consensus is that tools that are critical to many important advocacy organizations, community groups, and voter engagement groups are likely to decline in their reliability and utility, starting now in 2024.

Core Challenges

One fundamental challenge with the field of progressive tech infrastructure are the unique qualities of this corner of the larger tech market. While there are many consumer- or business-facing products that organizations use for a variety of core functions (like JustWorks, Salesforce Hubspot, Stripe, Square, and Slack), there are a host of activities that are quite specific to the political/civic engagement arena that are not serviced by such tools.

One hard truth about this sector is that it’s relatively small and not likely to get much bigger, since the number of elections doesn’t vary much and most advocacy organizations operate at a steady state. The kind of fast or exponential growth that excites venture capitalists doesn’t happen here. While every presidential cycle is a driver of investment and innovation, that’s not a reliable source of ongoing income. Government is a far bigger buyer of information technology than non-profit advocacy organizations, and far more investment has flowed toward gov-tech than political- or civic-tech since the turn of the 21st century.

Finally, like the larger tech industry, political tech changes rapidly, as do the products that consumers use and the behaviors that they adopt as they use those products. Unfortunately, the relative lack of investment in political tech doesn’t matter to a campaign volunteer who doesn’t understand why the canvassing tool she is trying to use on her phone takes seconds to update, or has a confusing interface, or drains her battery too quickly. And as soon as the major political tech providers manage to catch up to industry trends, tech takes another leap. (And no, I didn’t ask ChatGPT to write this paper.)

The Innovation vs Protection Paradox

One way that Democrats and progressives have navigated these challenges since 2007 has been to give a tech solutions vendor like NGP VAN a de facto monopoly position in the field. When the DNC and the Association of State Democratic Parties picked that company to be their sole provider of software for managing how they, and their state and local candidates, would access the voter file, a standard was set that not only allowed a host of actors to learn how to make best use of the NGP VAN toolset, the company also benefited from not worrying about being suddenly displaced by competition.

This protected position was very good for NGP VAN as well as for the national and state Democratic parties and political professionals. Since the DNC voter file was continuously enriched with fresh data as campaigns engaged with voters, that file—and access to it—got even more valuable over time. Knowing how specific voters in a particular district had responded to various messages or candidates is gold for campaigns. And for campaign professionals, once you knew how to use VAN’s tools, you had a necessary and portable skill that you could take wherever you went in the politics business.

But at the same time, being a protected monopoly also gave NGP VAN the freedom to innovate less.[5] For many years, NGP VAN took advantage of this privileged position to lock out competitors by denying campaigns access to its application programming interface (API) if they were also using any tools they deemed competitive. As Brian Young, the founder and CEO of ActionNetwork, an online organizing platform built in partnership with the labor movement, told me, “Their API never really worked to integrate with other tools very well. So, it disadvantaged tools like us in that space.” He added, “The fact that VAN was kind of the de facto CRM database of our movement has limited a lot of the ways that we can approach advocacy and building power, because it's all around an extremely inflexible limited data model of people.”

It also doesn’t help that the primary buyers of core political tech infrastructure who in theory are in the driver’s seat when it comes to demanding quality and accountability from tech and data providers—the national party and the state parties—themselves experience very high levels of leadership turnover. Since the DNC signed its first contract with NGP VAN, it has had eight CTOs, roughly one new one every two years.[6] State parties are also thinly staffed and have had little support, until recently, handling basic tasks like maintaining and upgrading their core tech infrastructure (kudos to Martha Laning of the State Parties Advancement Network and Nicole Aro of STAC-Labs, who work closely with dozens of state parties on these challenges).

But innovation is happening. A new generation of tech leaders who have cut their teeth working with first-generation tools like VAN are now deploying modern alternatives. Says Ari Trujillo-John, a twenty-year movement veteran who is the founder of OpenField, one of those alternatives, “VAN is very good at turning out people to vote who are registered to vote and who are probably already willing to vote.” But that’s not enough, she argues. “It is not good at helping us to earn the right to organize folks like [civil rights organizer] Bob Moses would have said. It's not good at building over time, expanding the electorate, expanding the base, or motivating people on issues.”

What is to be done?

Part Three of Living with VANxiety explores what ought to be done about the current situation with progressive movement tech. Here are my top findings and conclusions:

Over the course of this research, I have encountered a variety of perspectives. At one end of the spectrum are people who believe it is vital now to somehow claw back NGP VAN and ActionKit from Bonterra, by raising a fairly large sum of money and using an existing or new entity that could own and operate them on behalf of the whole sector. On the other end are people who believe there are already a slate of workable alternatives and that it is about time everyone stopped relying on NGP VAN’s relatively dated software suite. In between are people who don’t think it’s financially feasible to pry these platforms away from Bonterra, but who also think that there are important gaps in the current infrastructure that need to be addressed, and that ownership and governance are key questions to elevate alongside sustainability.

One thing everyone seems to agree about: despite the billions that the party committees, campaigns, and advocacy organizations spend on paid advertising and digital media, there isn’t enough money flowing in a consistent way to shore up movement tech’s core infrastructure. But even on that topic, disagreement arises; some believe the “incubator-accelerator” model of drawing private capital towards innovative political tech startups is succeeding as a way of building the next generation of tools, while others say that model is an inherently bad fit for developing and sustaining movement tech.

First, a bit of good news. The challenges presented by the Bonterra takeover are not immediately life-threatening to most advocacy and movement organizations. For example, in December, State Voices issued its recommendations for the tools and technologies that its affiliated organizations could use for the 2024 election cycle and most of them are outside the immediate orbit of Bonterra.[7] Most important, for organizations that need to access, use and store voter data, State Voices recommended two tools that could well make NGP VAN irrelevant: OpenField and CTA PAD. OpenField is a distributed canvassing app similar to MiniVAN, the backbone for campaign field operations, but designed to do more than just turn out people who don’t move frequently and are already registered to vote. And CTA PAD is a centralized data warehouse built by the Community Tech Alliance that groups can use to securely store and manage data, another key function provided by NGP VAN’s Votebuilder.

Those two functionalities—a tool that makes it easy for organizations to collect and manipulate data about individuals to facilitate outreach and manage relationships, and a data warehouse for securely storing that information and potentially sharing it across organizations—are at the heart of the current dependency on NGP VAN. (Access to a master voter data file is provided by either TargetSmart or Catalist, rival companies that were not directly affected by the Bonterra roll-up.) Again and again, what I heard from experts in the middle of the messy transition now underway in progressive tech is that if the movement can develop and sustain alternatives that provide those core functions and if they do so in an interoperable fashion, it will be able to weather any decline of NGP VAN.

Arthur Thompson, the DNC’s CTO since 2022, told me, “The Democratic Party has a state-of-the-art data infrastructure that allows us to reach our voters, provides critical information to campaigns and state parties, and sets us up to win from the local to the national level year after year. Our data and tools are some of our strongest assets for campaigning, so we bake in redundancies and take careful measures to ensure they are secure and resilient to any potential risks." Indeed, the DNC has an in-house large technology department which has invested heavily in recent years in reducing risks and dependencies in its data and tool portfolios and in developing its own capabilities for expressly partisan uses and storage of voter data. But the DNC is a cash machine. The same is not true for tech organizations focused on the non-profit side of the progressive ecosystem.

Interoperability, which means that key players all provide easy-to-use open APIs to their systems, is central to this future vision. No one wants to see a repeat of the 2007-2023 monopoly model of tech infrastructure that the DNC took with NGP VAN. And indeed, late last year, the Association of State Democratic Committees quietly ended its insistence that campaigns only use VAN to access their voter files. Rising alternatives include OpenField, Universe (which focuses on down-ballot campaigns), Civitech (which offers a “campaign in the box” suite of modern tools) and Votivate (which was developed by an affiliate of the Working Families Party to offer to endorsed candidates who were locked out of VAN access by state parties, and which is now launching an AI-powered array of next-generation tools for campaign managers and field operations[8]).

There may also be positive news on the horizon for organizations dependent on ActionKit. Demtech.ai, a new movement tech development venture based in India, is close to deploying an ActionKit alternative it is calling BeCause. Jhatka.org, a sister organization to MoveOn that was launched in India in 2012, will be its first user. As Avijit Michael, a founder of Jhatka who is now leading the Demtech.ai effort, told me, ActionKit has long been an extremely expensive proprietary option for organizations based in the developing world. “Why don’t we develop here?” he asks, noting that the cost of top software development talent in India is less than one-tenth what it costs in the United States. Assuming the first version of BeCause succeeds, Michael says that MoveOn is considering investing in the startup to assure that version 2 will be designed to meet its needs. This could be a major boon for the entire advocacy sector in the US.

Still, there are thousands of campaigns and organizations that are reliant on Bonterra’s offerings, and the cost of training tens of thousands of staff and millions of volunteers on new and unfamiliar tools is seen as prohibitive by some. NGP VAN is also battle-tested and its code is unlikely to fail in the final weeks of the election cycle, something no other new platform can promise. And because of its long-time dominant position in the field, NGP VAN’s choices about how to define particular kinds of data and metrics function as de facto standards; a switch to a more competitive landscape means that what one provider defines as a “door-knock” may not match precisely with what another provider says. These are the main reasons why one school of thought insists that somehow, the “movement” should buy NGP VAN from Apax. An optimistic estimate of what this would cost is $50 million, based on declining revenue expectations, one expert suggested. Others thought that Apax wouldn’t sell for anything less than $100 million. No one offered first-hand knowledge of what the overseas hedge firm really wants were it to unload NGP VAN.

Given that general understanding that this long-serving toolset may be approaching obsolescence, the words of Julia Barnes, the CEO of The Movement Cooperative, are worth hearing in full:

"If we have the chance, I would be in the camp of ‘the movement needs to buy back NGP VAN’ because even if it is an archaic codebase, and they have stopped meaningful development on a lot of the tools that are actually good, like ActionKit, it's better we own it than somebody else. And if we're going to put this thing out to pasture, we should do it on our own terms. We should benefit from the 20 years of support, development, and learning that we have all funded for that company. And we should build the next thing or the next set of things from a place where our data and our trajectory off of that platform is protected. I would also add that we have an entire generation of data staffers that is aligned with them and they have done us good service and deserve to continue their expertise divorced from the politics of ever-increasing profit demands.”

The Movement Cooperative is a centralized data infrastructure and technology hub for the movement made up of about one hundred progressive groups that share costs for software tools and data. Barnes adds, “I am interested from the perspective of The Movement Cooperative in finding a vehicle or vehicles that can be funded through zero return philanthropic investment or direct investment from movement partners, similar similarly to a co-op, where the movement venture fund that that would create has the opportunity to build and invent the core infrastructure from a nonprofit model, so that we never have to question the financial viability of our most essential tools.”

What is the core infrastructure that such a co-op would support? Barnes rattles the answer off easily. “A data warehouse, a voter file, a CRM, a canvassing functionality, a phone functionality, an email functionality. We have to have all of that, and it has to be in designed in a way where it has an open API, where it can talk to all of the tools that people want to sync in, where the users get to determine what aspects of their own data are important, as opposed to that being designed by the company.”

Robin-hooding from billionaires, or?

Can this be done outside of the for-profit model of tech development that has mostly shaped and driven the field? Barnes doesn’t equivocate. In her experience, leaving tech development to the handful of six or seven wealthy individual investors who currently drive Democratic tech feeds an ultimately destructive cycle. Startups get seed investment that carries them through their first two- or four-year political cycle, and they scale quickly to prove their viability. Then, both to satisfy their initial investors and to keep attracting other capital, they either must apply the competencies they have demonstrated in the field of politics to the larger market, or they have to acquire other functionalities to keep expanding, or they need to find a buyer. A movement cooperative that owns, sustains, and develops core tech infrastructure, Barnes argues, “is the only way that this doesn't become a cycle every ten or fifteen years for the movement where we have to wrench back control from the for-profit capitalist approach that has been incubated in our incredibly small and never-growing market.”

Right now, there are basically two ways that progressives are sustaining existing shared tech infrastructure and capacities. The first is what one veteran technologist I spoke to jokingly referred to as “Robin-hooding from billionaires.” Benevolent donor-investors like Reid Hoffman, Eric Schmidt and Laurene Powell Jobs along with the Mind the Gap donor network have done the lion’s share of this kind of funding. In the immediate aftermath of the 2016 election, when anti-Trump passion ran high and a lot of people were sensitized by Hillary Clinton’s loss to tech deficits allegedly or actually faced by progressives, money flowed easily. Now it is flowing more fitfully.

The other alternative is exemplified by the ownership and management models that have carried ActionNetwork and Catalist for more than ten years. Both are hybrids, for-profit businesses owned and governed by nonprofits with boards made up of movement leaders. [See Part Three of the report for the gory details.] OpenField, which is a much younger company, has a similar model, with a group of movement investors sitting on the board of a C corporation who have the explicit power to veto any sale, and a platform development committee who the company is accountable to for development decisions. These constraints don’t interfere with these entities finding a viable business model, but they have kept these organizations’ leaders from pursuing the always-seductive opportunities of profiting from serving the much larger private sector.

Conclusion: Towards Sustainable and Secure Movement Tech

Organizers, activists and the institutions that house them and channel their energies need technologies that they can count on to be reliable, versatile, affordable and secure. While the larger commercial tech sector will always meet some of those needs (and come with its own risks–see Elon Musk’s purchase and debasement of the “digital public square” that was Twitter), the civic and political engagement field has unique requirements. Unfortunately, that field is not now and probably will never be big enough to justify large-scale capital investment. Besides, as we have seen, the venture-capital model for early-stage companies in the movement tech field either creates unrealizable expectations and pressures for fast growth or drives these companies away from their core missions toward the needs of the larger private market for tech tools.

Expectations created by the larger commercial tech sector also deform the field of movement tech in one more subtle way. Users of tools made by companies like Meta, Google, Twitter/X, Amazon et al have come to expect to pay nothing for those services, not realizing that when an online product is free that is because they, the users, are the product being exploited. The online advertising market is worth tens of billions to these companies. And while movement tech is centered on helping organizers and activists make smart use of individual data, these organizations often promise to keep user data private and ethically shouldn’t get into the business of exploiting their users’ data by selling it to third parties.

Millions of individuals rely on movement technologies on a daily basis, but few have any understanding of what it costs to build, maintain and extend them. So how shall we pay for movement tech? 

One idea that kept arising over the course of my research was the notion of a movement tech trust, set up and governed by key movement stakeholders, that would support core infrastructure. “Just two percent of the $1.5 billion that is going to be spent on paid media for the 2024 presidential election would sustain us for three or four years,” Kat Atwater, the founder of the Community Tech Alliance, a provider of data infrastructure, told me. But there’s no obvious or reliable way to get the mega-donors who put up the lion’s share of that money to set aside a small portion for tech. And no one has yet found a way to get the millions of small donors who now contribute billions to political and civic engagement organizations and campaigns to also support movement tech.

But imagine if every time someone donated via ActBlue, a box popped up that offered them the option of adding a few dollars to a Movement Tech Trust Fund. Or, whenever a volunteer finished a canvassing shift, the app on their phone thanked them for their time and then asked for a small contribution towards the upkeep and development of similar tools. Or likewise, at the end of a virtual phone bank, a voice message played reminding the volunteer that the tool they just used was supported by the Movement Tech Trust and asked for a small gift. In 2020, more than 15 million individual donors gave money to campaigns and organizations via ActBlue. Millions volunteered to canvass and phone-bank. So, at least in theory, there’s a resource base to be tapped.

It’s time to develop a different model for funding movement tech. Otherwise, what happened with NGP VAN and Bonterra won’t just be a cautionary tale from the past. It will be a harbinger of the future.

---

[1] https://www.businesswire.com/news/home/20210602005376/en/EveryAction-Makes-6th-and-Largest-Acquisition-with-Salsa-Labs.

[2] It’s a safe guess that Apax valued EveryAction at roughly 5x its annual revenue, which would imply a purchase price of $500 million.

[3] Liana Baker, “Apax Merges Software Firms for Non-Profits in $2 Billion Deal,” Bloomberg, August 3, 2021, https://www.bloomberg.com/news/articles/2021-08-03/apax-merges-software-firms-for-non-profits-in-2-billion-deal

[4] Trevelyan announced his retirement “from 26.2 years” in the field at the end of November 2021 https://twitter.com/stutrev/status/1465401782859927559.

[5] Will Conway, “Why the NGP VAN model hurts democracy,” Medium, December 21, 2015. Conway, who was then working for NationBuilder, a non-partisan competitor to NGP VAN, connected a serious security breach that led to the Bernie Sanders campaign accessing proprietary information belonging to the Hillary Clinton campaign to NGP VAN’s insulation from competition. He wrote, “They have a clear hold on their focus market, Democratic politics, so they have absolutely no incentive to improve their product.”

[6] Ben Self (2005-2009), Josh Hendler (2009-2011), Bryan Whitaker (2011-2013), Andrew Brown (2013-2017), Raffi Krikorian (2017-2019), Lindsey Schuh-Cortes (acting CTO 2019), Nellwyn Thomas (2019-2022), Arthur Thompson (2022-present). To be precise, the role was only elevated to “Chief Technology Officer” in 2017.

[7] See https://drive.google.com/file/d/1652hYUQkS-iGlWwbwYdklEr2L9neo4V9/view

[8] See “AI is on the verge of changing the practice of campaigning itself,” https://theconnector.substack.com/p/ai-is-on-the-verge-of-changing-the (The Connector, January 24, 2024).