Understanding the impact of vote.gov on the public during the 2024 election year

Challenge
With over 11 million users during the 2024 election year, vote.gov needed to understand whether its newly redesigned platform was effectively serving all eligible voters—particularly those facing language, accessibility, and situational barriers. The site had launched major improvements mandated by Executive Order 14019, but lacked comprehensive data on whether these investments were actually making a difference for the American public.

Client
vote.gov, U.S. General Services Administration (GSA)
The federal government's official voting information resource, serving as a non-partisan aggregator directing citizens to state-specific election information.

Contribution
Lead UX Researcher & Impact Analyst
I designed and executed a multi-method research strategy combining analytics, population data, stakeholder interviews, and user testing to evaluate impact across six primary goals. I also consulted on the July 2024 site launch, conducting user testing, optimizing search functionality, and advising on visual design decisions to ensure consistency and representation.

Impact

• Validated 9,000-10,000 voter registrations attributable to vote.gov during September 2024 debate traffic surge through statistical modeling

• Demonstrated proportionate reach across demographics: non-English content reached 4.4% of users vs. 5.7% LEP population

• Confirmed 53% completion rate for National Voter Registration Form tool—exceeding industry benchmarks

• Established baseline metrics for measuring accessibility impact across 40+ million eligible voters with disabilities

• Improved search experience by cleaning up outdated results and surfacing high-priority content based on user query analysis

• Enhanced navigation and content based on insights from users who struggled to complete tasks on the site

Key findings

State referrals and impact attribution

Challenge: Vote.gov doesn't register voters directly—it refers them to states. How do we measure success when the moment of conversion happens on another platform?

Approach: When Taylor Swift shared a link to vote.gov following the September 10 presidential debate, it created a natural experiment. The site handled 500,000+ requests in five minutes without downtime. Because her referral link was uniquely trackable, I saw an opportunity to measure impact.

Working with statistician Jake Bowers, I requested registration data from states with online registration. Four states—Colorado, Rhode Island, South Carolina, and Oregon—shared their numbers. We compared users who clicked from vote.gov to each state's website against the number of online registrations recorded, calculating the maximum possible registrations attributable to vote.gov as the lesser of those two numbers.

Using the week before the debate (September 3-9) as baseline, we modeled expected registration numbers for September 10-11, then compared predictions against actual registrations.

Key findings:

• Identified 9,000-10,000 registrations beyond baseline expectations across four states

• Swift's unique referral link accounted for 34% of state site clicks, attributing 3,000-3,400 registrations specifically to her endorsement

• Demonstrated vote.gov's value as an authoritative aggregator that amplifies influencer/celebrity impact

• 65% overall click-through rate from vote.gov state pages to state election websites

• Consistent 1-2% reach across all states and territories relative to eligible voter population

Strategic insight: This analysis proved vote.gov's impact despite structural tracking limitations. The path forward requires state partnerships with ongoing referral tracking through data-sharing agreements, plus clearer messaging about vote.gov's aggregator role to manage user expectations.

"There were many languages listed and it's impressive. It shows the degree that government is seriously investing in including LEP people. Recently, for instance, my parents received ballot information in Korean. It shows government really strives for inclusion."
— Participant in Korean-language study

Language access and translation quality

Challenge: Vote.gov invested heavily in human translation for 20 languages covering 97% of the American population. Was the investment reaching its intended audience, and did translation quality affect trust?

Approach

• Analyzed traffic-to-population ratios across all languages using Census data

• Conducted UserTesting studies with 24 bilingual participants rating translation quality and impact on trust

• Interviewed language access advocacy organizations about community needs and dissemination strategies

• Analyzed user session data to understand discovery patterns

Key findings

• Non-English pages reached 4.4% of total users, aligning proportionally with 5.7% LEP eligible voter population

• Ratio of users to population was consistent across most languages at 1-3%

• Participants rated vote.gov translations 6.1/7 for quality vs. 5.6/7 for government sites generally

• Multiple participants explicitly connected high-quality translations to increased trust in both content and government

• 74-84% of non-English users arrived at vote.gov already in their preferred language via search engines

• "Voting as a new U.S. citizen" guide was most-visited guide in 11 of 18 non-English languages

Strategic insight: The investment in human translation paid off in both reach and trust, but impact could be amplified significantly through partnerships with ethnic media and community-based organizations rather than relying solely on organic search discovery. Advocacy groups emphasized that trusted messengers are essential for reaching LEP communities.

Accessibility and universal design

Challenge: How do you evaluate accessibility impact when you can't track assistive technology use, and the population of people with disabilities is highly diverse?

Approach

• Conducted objective accessibility audit against WCAG 2.1 Level AA standards and Section 508 requirements

• Gathered feedback from accessibility advocacy organizations who shared vote.gov with their communities

• Ran UserTesting study with ADHD participants (N=9) on website usability

• Evaluated National Voter Registration Form tool completion rates and user feedback

• Consulted on design decisions throughout site development

Key findings

• Vote.gov met or exceeded all accessibility requirements with no significant violations

• NVRF form tool achieved 53% completion rate—exceptionally high for government forms; the tool made 27-page PDF accessible on mobile devices and with screen readers

• Users with disabilities validated that prioritized features (plain language, logical organization, zoom capabilities, clear navigation) genuinely improved experience; more than half of accessibility advocacy respondents noted thoughtful features enhanced trust

• ADHD participants praised calm design, absence of distractions, and clear information hierarchy

• Estimated 40+ million eligible voters with disabilities potentially benefited from improvements

Strategic insight: Meeting standards is essential, but providing truly accessible experiences requires ongoing testing with users with disabilities and continued partnership with advocacy community—both for improvement feedback and amplified dissemination.

"Very clearly organized and helpful links embedded in the document. Also answers very key questions that many expats will ask."
— Research participant, Voting as a U.S. citizen living outside the U.S.

"I like the structure of it and I think it has just the right level of detail for people who might need help with this."
— Research participant, Voting as a military service member

"It gives me a level of confidence and an awareness of what, to know, what to expect, and how I can actually defend myself as and when needed."
— Research participants, Voting as a new U.S. citizen

Voter guides for situational barriers

Challenge: Beyond language and disability, could vote.gov effectively serve voters facing specific life circumstances that impact their ability to participate?

Approach

• Created eight voter guides based on primary and secondary research with target populations

• Content tested guides with representative users, measuring usefulness and gathering qualitative feedback

• Analyzed traffic patterns and compared to estimated target population sizes

• Examined language preference patterns to understand which guides served multilingual audiences

Key findings

• Content testing participants rated guides 4.5/5 for usefulness on average

• Guide traffic represented 12% of overall site traffic (August-November)

• "Voting as a college student" guide had highest overall traffic with 404,667 users (~3% of target population)

• Strong correlation between user demographics and guide usage—particularly for new citizens guide in non-English languages

• Positive feedback highlighted clear organization, useful links, and consolidated information from multiple agencies

Strategic insight: Vote.gov filled critical content gaps that no other voter information site addressed comprehensively. The opportunity lies in dissemination—partnering with organizations serving these populations (colleges, naturalization programs, reentry services, housing assistance) to ensure guides reach people who need them.

Search optimization and site improvements

Challenge: Users struggled to find information they needed, and search results surfaced outdated or irrelevant content.

Approach

• Implemented Touchpoints feedback form and recruited 10 users who had struggled to complete tasks on vote.gov

• Monitored search.gov queries in English and Spanish to understand common questions

• Systematically cleaned up outdated prioritized search results

• Analyzed search patterns to identify content gaps and homepage visibility needs

Key findings

• Direct user interviews revealed specific navigation pain points and content gaps

• Search query analysis identified frequently-asked questions that weren't prominently addressed

• Bilingual search patterns revealed similar information needs across languages

• Outdated search results were creating friction and poor user experience

Strategic insight: Ongoing search monitoring and direct user recruitment provide essential feedback loops for continuous improvement. Small optimizations based on real user behavior compound into significantly better experiences.

An email exchange with one of my research participants after the session. From our session, we learned that concerns about tax implications of registering to vote from abroad had stopped this voter from registering and exercising their right to vote for 17 years. This demonstrated how much you can learn from talking to real users and reminded me how much impact every interaction the public has with the government or a representative of the government can have on trust and perception.