What I learned running digital organizing campaigns in rural Oregon

Peter Marks
9 min readMay 30, 2021


Me walking off into the sunset in Tillamook County, feeling half-victorious and also cold.

I moved to Tillamook County, Oregon at the beginning of the pandemic while I was working a series of nationally-focused political technology gigs leading up to the 2020 general election. In the weeks leading up to that election I got involved in local campaigns in Tillamook and have continued to stay involved. In the time I’ve spent since organizing on a volunteer basis, I feel I’ve learned as much about organizing as a paid staffer on national campaigns. I wrote this article to share what I learned running digital organizing programs in this area in hopes that it helps organizers running campaigns in similar areas.

For some context on Tillamook County, it’s an incredibly beautiful rural area on the North Oregon Coast. It’s best known for the Tillamook Dairy co-op, the county’s biggest employer that exports its cheddar cheese to seemingly every pocket of the world. There are around 20,000 registered voters with fairly even numbers of registered Democrats and Republicans, but the county has voted more conservatively in recent years. It voted twice for Obama and then twice for Trump. I find it to be a generally pleasant place to live filled with nice people, but I’m not sure I’d feel the same way if I wasn’t white. Like a lot of Oregon, there’s a history of white supremacy in this part of the state, which is documented in a book called A Brief History of Fear and Intolerance in Tillamook County.

The campaigns I supported were part of a joint effort called the Campaign School that focused on the May 2021 special district elections. The effort was a collaboration between The Tillamook County Democrats and Tillamook Vote Forward, an Indivisible and Swing Left group led by an experienced politician, Beverly Stein. Beverly led the program and was an invaluable resource that we all learned a lot from, myself included. The goal of the effort was to recruit first-time progressive candidates and help them run their digital organizing programs. Campaign School supported seven candidates, four of whom won. My personal focus during this effort, and the focus of this article, was on the following four contested campaigns:

  • Jaime Perez, Garibaldi Port Board Commission Position #2 — Jaime was a young business owner and first-time candidate. He entered a race against two more traditional candidates and won by 47 votes. My personal highlight of the campaign was getting a celebratory drink with Jaime and his campaign manager the weekend after the election and hearing that they were still stunned by the result as they thought they had no chance of winning.
  • Mary Johnson, Transportation District Position #5 — Mary was another young, first-time candidate who was the only candidate I work for who ran for county-wide office. She ran against a very experienced candidate and won by less than 500 votes with 54% of the vote.
  • Constance Shimek, North County Recreation District Position #5 — Constance ran a competitive race against a tough incumbent. While she did not emerge victorious, she made an impressive commitment to stay positive in the face of negative campaigning and wasn’t afraid to innovate with new tactics.
  • Andrea Goss, Tillamook School Board Position #1 — From my perspective, Andrea and her campaign manager Laura Rochelois ran the most impressive campaign of anyone we worked with. Andrea personally called thousands of people in a very tough district. While she did not emerge victorious, her candidacy raised hugely important issues in the community and was a much needed voice amongst a slate of conservative candidates spreading fear about Critical Race Theory in schools. This loss hurt and sadly it may impact subjects students in Tillamook County are allowed learn.

A lot of effort that went into these campaigns, including mailing postcards, preparing candidates for community forums, fundraising, creating a social media presence, etc. My role on these campaigns and the focus of this article is what I learned running digital organizing campaigns, primarily through NGP VAN, the primary software used for Democratic campaigns. Internally, I was referred to as our group’s “VAN Czar”, a title I fully embraced. We expected a relatively low turnout as is the case with most local non-partisan elections, so we felt these tactics had a large potential for impact. It should be mentioned that this election was of course run during the COVID-19 pandemic, which limited a lot of tactics such as door knocking. Here are my main takeaways from the experience:

1. Relational organizing helped recruit candidates

Our first challenge of the Campaign School was finding candidates to run. We didn’t exactly want to broadcast our strategy by putting an ad out online, so we asked our friends, neighbors and contacts within community organizations if they were interested in running or knew of anyone interested in running. This is what some people describe as “relational organizing” and others would describe plainly as “organizing”. One criticism of relational organizing is the perception that you are limited by your personal network, but I personally don’t think it has to be that way. One thing I try to teach and others picked up on is that you should always be asking your contacts “who can you introduce me to who knows people”. Strangers are far more interested if you preface the conversation by saying “[person who you have a relationship with] asked me to call you”, which leverages your relationship.

We got several candidate prospects this way, but we found they often needed someone like Beverly to close them on the idea of running. Having helped to design a free-to-use relational organizing app at the Organizing Empowerment Project called Empower, I convinced folks we use it as a recruitment tool. While I believe that app is useful for a number of organizing situations, I wouldn’t recommend introducing it in this situation and would just recommend people track leads in a spreadsheet.

2. Phone banking without an auto dialer was a slog

With broad advisories against knocking on doors during this moment in the pandemic, phone banking seemed to be the best way to connect with voters we had no relation to. Our campaigns used the “virtual phone bank” feature in VAN that requires the caller to use their own phone to manually call people. This is different from the “auto dialer” tools like ThruTalk I’d used on larger campaigns where the software attempts multiple calls at once and connects you with whoever picks up. By comparison, VAN’s “virtual phone bank” tool is a slog. As a result only one of our candidates was able to make time for it. Impressively, Andrea Goss made thousands of calls this way and attempted to reach just over half of the ~6,000 registered voters in her district. She canvassed about 10% of total voters via phone, which I’m sure was impactful but required a lot of time. Of the other candidates, two of them canvassed less than 1% of their voter universe via phone and the other one did not phone bank.

I remain convinced that phone calls are an integral part of voter contact and are not to be dismissed. However, I’m not convinced that manual “virtual phone banks” are worth the effort. Next time, I would advocate using an auto dialer and potentially running a calling campaign for smaller races as a centralized effort. We ran a similar centralized effort for text-banking that efficiently recruited volunteers through a multi-shift Mobilize event.

3. Text-banking was integral to building name recognition

Since there was no official voters pamphlet for this election in Tillamook County and we were running first-time candidates, building their name recognition was of the utmost importance. “Peer to peer” text messaging, a semi-automated way of sending bulk text messages without using a personal phone, was a tactic that helped us do that and more. Studies show that while voters will often delete an unsolicited voicemail before they listen to it, they usually read an unsolicited text message even if they end up replying “STOP”. Given the reach of these campaigns and the high likelihood that they were seen, this tactic probably helped with name recognition more than any other tactic we used.

Our campaigns delivered 14,123 text messages to get out the vote, more than 20x the number of calls completed. This was done with about 20 volunteer hours and a combined total of about $400 through a low cost service called Scale To Win that I highly recommend for both their product and support. Each candidate had their own Scale To Win account hooked up to their VAN account, which allowed them to text to lists in VAN and have survey questions ported back to VAN. Individual accounts meant that unsubscribing from one campaign would not unsubscribe you from other campaigns, which gave campaigns a high-level of independence and autonomy in their texting. Most campaigns sent text messages to every group of voters (Registered Democrats, Registered Republicans and non-affiliated voters), each with a customized message. For example, we identified our candidates as Democrats to registered Democrats but not to other groups.

There was a fair amount of hesitance about this tactic as some people in our Campaign School group dislike receiving unsolicited political text messages. However, the response totals painted a largely positive picture of the experience. 57% of the replies ended up in support of the candidate, 6% ended up opposing the candidate, 7% were left undecided on the candidate, 3% already voted and 25% indicated we had contacted the wrong number. On a personal anecdotal note, I had some truly memorable exchanges with folks that I learned a lot from and persuaded countless voters to support our candidates in what felt like a matter of minutes. Setting up text message programs felt like the highest value thing I did for our candidates and I would recommend it to anyone running a local campaign. In the words of our contact at the Democratic Party of Oregon, “do texting now before the government regulates them into oblivion”.

One of the things I suspect we could have done better with our texting campaigns was to set them up in such a way to learn from the different kinds of messaging. I advised most of our candidates to keep it short and simple, but a few tried some promising new messaging approaches. Given that they were all messaging different demographics of people in different districts as different campaigns, it’s really hard to compare the numbers in a fair manner. Anecdotally, the messages coming from candidates themselves that looked conversational saw the highest rate of engagement. Given the response rates, I’m not sure it’s possible for campaigns reaching out to less than 5,000 voters to get statistically significant data. County parties or other coalition groups of local candidates are probably better positioned to learn here. It’s worth mentioning Tillamook Democrats did their own texting efforts for GOTV, but I cannot speak to the specifics of that effort.


I found working on these campaigns quite meaningful and uplifting in these strange, isolated times. I’m incredibly grateful to all the candidates who put their names on the line for our community and all the volunteers who spent their weekends reaching out on our candidate’s behalf. Local elected officials play hugely impactful roles, especially in this environment of increased right-wing extremism. Given the historically low-turnout in local elections with persuadable voters, I feel that supporting these campaigns is some of the highest value work you can do as an organizer.

Reflecting on this experience through the lens of someone who works full-time in political technology, I have a new appreciation for the amazing array of features that NGP-VAN provides. And yet, I see how it alone wasn’t enough and that additional tools were needed to scale voter outreach. It’s asking a lot of first-time candidates and campaign managers to learn to use this software and even more to ask them to extend it with third-party outreach tools. I can’t help but wonder if votes are being left on the table because of VAN’s learning curve, but I hope this article helps other first time candidates and first time “VAN Czars” get the most out of these powerful digital organizing tools.

The results of this election also make me confront the reality that perhaps digital outreach isn’t everything. We felt like we were out-organized in one of the races we lost through well-established and mostly offline networks like churches and farm bureaus. This is a healthy reminder of what I’ve been hearing from experienced organizers: that phone calls and text messages sent in the weeks leading up to the election are “mobilization” tactics, but not necessarily true organizing that happens in the context of real relationships in our community. Something to keep in mind for the next round.

Special thanks for this program is due to Jonathan Rudolf and Ron Morgan of the Democratic Party of Oregon for their VAN training sessions and support. Special thanks also due to The Tillamook County Democrats for making VAN affordable for our candidates.



Peter Marks

Co-founder of Oregon Rebate 2024 campaign. Former Presidential Innovation Fellow and tech staffer of the Bernie Sanders 2020 presidential campaign.