Opportunities in Campaign Technology
In the 2018 midterms, DigiDems embedded over 80 technical workers in a wide variety of campaigns across the country - congressionals, gubernatorial and state-wides. This has led to a wealth of knowledge, generated from the ground up, uncovering the most pressing gaps in the campaign ecosystem. This is a synthesis of the spaces where our organizers observed a need that did not have a suitable solution.
Our DigiDems organizers were instructed to add value by identifying spaces where they could use technology to help their campaigns win. Most worked in field, finance, or social media -- a skew which informs our perspective.
These observations were synthesized from:
a series of individual interviews with each DigiDem
retrospective documents that prompted our DigiDems to reflect on their experiences
a conference after the election where knowledge was shared and captured
common themes found in the questions asked in the online forum
a GitLab repository that houses the tech tools that were built in response to gaps in their campaigns
By nature of where our DigiDems existed on their campaigns and their bottom-up perspective, this is not an exhaustive list of opportunities in the political tech space. For example, there are certainly opportunities in advertising, operations, opposition research and more, yet our observations from the ground do not speak to these gaps.
We’ve noticed the difference between a good and a great tool tends to be these characteristics:
Flexible and Customizable: Each campaign is unique and must retain enough control over the tool to be able to tailor it to their needs.
Well-Documented and User Friendly: Time is one of the most valuable resources on a campaign, and a tool that is hard to understand immediately will likely be discarded.
Built with Strong Version Control: Many people may be given access to a tool without a strong understanding of how to use it. Being able to undo mistakes is essential.
Scalable: As a campaign builds, stress placed on tools will gradually increase until GOTV. Tools must be able to accomodate all phases of the campaign cycle, and can not break or slow down when they are needed the most.
Integrated: A new tool must communicate with the existing softwares that are widely used or risk being a burden for the campaign to piece into their workflows.
Secure by Default: Security best-practices are enabled and enforced by default. These include security-key based 2FA for user accounts; data encryption at rest and in transit; role based tiered access to data; user auditing; ability to back-out maliciously input data (troll-removal).
Field remains a cornerstone of the modern campaign, yet teams continue to struggle with measuring the impact of their field strategy.
Field teams need better ways of drawing realtime, actionable insights about the success or failure of contacting and persuading voters in their district. Existing measures of success on a campaign are centered around doors knocked and calls made, but do not quantify or visualize the effect of these actions.
We are interested in supporting technologists working on discovering new measures of success, and on tools that result in instantaneous understanding of the direct outcomes of field team operations.
Online Volunteer Communities
Volunteers often return for the community that can be found within the walls of a campaign office. Strong friendships have been built in working together for a shared belief, and friendly competition between volunteers on number of contacts made is sometimes hard-coded in rudimentary leaderboards.
Gamifying the volunteer experience may increase engagement, and having a portal where training materials and campaign information are easily accessible would benefit both the campaign and the volunteers. Currently, volunteers stay up to date with new campaign information through mass emails, phone calls, and occasionally Facebook groups.
We would like to see mobile and web applications that centralize a campaign’s volunteer community and further research around features that would boost in-person engagement. Tools focusing on increasing the number of repeat shifts, by improving the volunteer experience, could have a significant force multiplier effect on campaign efficiency.
A great field organizer remembers their volunteers and internalizes a map of their relationships. For example, if a husband and wife always canvass together, there is no need to call both in order to confirm they will be coming into the office. If a meet and greet needs a host, an organizer remembers who the most well-connected volunteers are within their network in order to maximize attendance. This information tends to be held in a free-form note in VAN or exists only with an organizer -- but it is the key to identifying the central people who can exert the most social pressure in mobilizing the community.
We are interested in tools that streamline the mapping of volunteer relationships, helping organizers visualize their volunteer network and leverage it to its full potential.
The process of cutting turf for a volunteer to canvass is, especially in the beginning stages of a field program, a fairly random operation. Turf-cutters eyeball areas that are dense but have not yet been walked, but little thought is given to which areas have the most potential for persuasion based on density, demographic information, etc. The question “If you could send only one canvas group out, where should you send it?” is not one that has a simple answer on a typical campaign. Rudimentary formulas in google sheets multiplying the number of people in a precinct with how much it has been already canvassed provide a simple priority order. When a volunteer is bilingual, the field organizer is expected to know where in the district would make the most sense to send them.
Tools that optimize the turf-cutting process to find the neighborhood that provides the best potential for persuasion in the district based on some set of criteria that takes into account both the volunteer resources at hand and the state of the campaign would drastically multiply the effectiveness of canvassing operations.
The advent of mobile canvassing applications has reduced the time that a campaign spends on entering data from paper canvassing packets. However, many volunteers are still averse to using their phones or do not have access to a smartphone and thus still canvass on paper. Paper canvassing made up around 30% of canvassing in the 2018 midterms (source? came from dnc slides).
The manual data entry that comes with a paper packet takes time, and during the crucial period of GOTV, is often dropped completely. A solution that works for all volunteers while minimizing data entry is still missing. Proposed ideas to cut down on data entry have included machine readable paper forms and OCR technologies. We are interested in new systems that would integrate seamlessly into existing tools such as VAN.
One is hard-pressed to find a field office that does not have at least a few maps posted on its walls. From dividing turf amongst organizers to visualizing district penetration to locating ballots left to chase, a field program currently relies on tools that are either too technical for the average staffer or come with minimal functionality. Relevant shapefiles are difficult to locate and work with, and real-time sync with Google Sheets or VAN is non-existent. Campaign staff rely on Google Maps or, when a more technical staffer is on the team, Carto. However, there is no tool that works, straight out of the box, for campaigns to easily visualize the data they need on a map.
New tools that simplify the process of creating insightful maps for the average staffer are an under explored frontier.
A campaign fields questions from voters across many mediums -- through social media, in person, by email, etc. Often repetitive in nature, answering these questions often involves copy and pasting a talking point from a prepared document. However, due to the volume of requests, this task can monopolize staffers’ time.
Tools that allow for quicker ways of communicating talking points by enabling easier answering of frequently asked questions, while filtering those that need a personalized answer would help campaigns engage with more people and ensure a high response rate. Furthermore, some analysis of which issues are the most asked about and therefore more important to voters may help in surfacing which talking points need attention.
While the shape of the following problems were surfaced by our research, the depth of the problem and the possible solutions were not elucidated by our organizers’ experiences on the ground. Further research is needed in these areas to clarify the scope of the need.
Fundraising is the foundational brick supporting campaign operations. A candidate’s call time is typically one of the most managed and time-consuming activities in their schedule. Prospecting new donors is a tedious and inefficient process, generally done manually by finance staffers.
Data-driven approaches to predicting who will donate, how much, and the best contact method can dramatically save time and increase fundraising. For example, it would be more efficient to text small donors, but cannibalizing time from high-dollar prospects that need a phone call must be avoided. Existing tools also tend to be expensive for the average campaign.
Campaigns need better ways of digesting online public opinion, media coverage, and current events so that they can respond appropriately and in real-time.
In his beard lived three cardinals.