Case study
SmarterVote voter matchmaking app
Built a cross-platform quiz app that ranks every candidate on a voter's ballot by issue alignment.
Industry
Government & Civic
Key result
Cross-Platform
iOS, Android and web
Services
Government & Civic
The Challenge
The Challenge
Voters want to make informed choices, but researching every candidate in every race takes hours — hours most people don't have in the weeks and days leading up to an election. The result is a predictable pattern: most people only know the top of the ticket. The president. The senator. Maybe the governor. No simple tool existed to compare candidate stances simultaneously across local, state, and federal races, so matching personal positions to the right candidates meant researching each one independently.
Down-ballot races — school board, city council, local judges, ballot initiatives — go largely uninformed. These are the offices that most directly affect daily life: the quality of local schools, public safety policy, land use decisions, and the administration of justice in a community. Most voters still walk into the booth with little to no information about the candidates in those races. Smarter Vote saw an opportunity to close that gap with a simple quiz: answer a few questions about where you stand on key issues, and get back a ranked list of candidates who share your positions, across every race on your ballot.
The harder problem was building something voters would actually use. That meant mobile-first, because the majority of Americans access information on their phones, not their desktops, and the civic tools that existed were clunky, outdated, and desktop-only. It meant fast — voters aren't going to sit through a 40-question survey. And it meant reliable, because the worst time for an app to go down is the evening before election day, when millions of people are doing their final research. It also had to work across platforms: a civic tool that only runs on iPhone misses half the electorate.
The Solution
The Solution
SmarterVote was built using React Native, which let a single shared codebase run natively on both iOS and Android. The web app was built with Bootstrap and jQuery on a PHP backend, keeping the server layer simple, fast, and easy to maintain during election cycles when reliability is critical. All three platforms — iPhone, Android, and web — pull from the same data layer, so candidate information is consistent everywhere. Building native apps rather than a website alone brought push notifications for election day reminders, home screen shortcuts, and smoother touch interactions.
The matching logic collects a voter's stated positions across a set of issue questions, weighting and storing each answer. On the candidate side, the platform holds each candidate's documented positions on the same set of issues, sourced from public statements, voting records, and campaign materials. When the quiz is complete, the app calculates an alignment score for each candidate in each race — the percentage of issues on which the voter's position and the candidate's position agree — and ranks candidates by that score within every race on the voter's specific ballot.
That algorithm is intentionally transparent. It doesn't make value judgments about which issues matter more; that is left to the voter. It simply reflects the match between stated positions. SmarterVote itself was deployed on Amazon Lightsail, part of AWS, chosen for predictable pricing, easy vertical scaling, and the underlying reliability of AWS's global infrastructure. Election traffic is the definition of a spike — on a normal Tuesday in October the app might see a handful of sessions, and on election day it surges by orders of magnitude — so the app stays up through the spike and costs return to baseline afterward.
In use, a voter opens the app on their phone or navigates to the web version and enters their zip code or selects their state. The app loads the races on their specific ballot. They move through the quiz, tapping their position on each issue: agree, disagree, or neutral. The whole process takes about three to five minutes. At the end they see their matched candidates ranked by alignment score for every race, from US Senate down to local school board, and can tap any candidate to see which issues drove the match, share their results, or save them for the voting booth.
Results
Results
iOS + Android + Web
Platforms shipped
Single shared build
Codebase
3–5 minutes
Quiz completion time
Amazon Lightsail (AWS)
Infrastructure
Every race
Ballot coverage
Issue alignment score
Candidate ranking
Before & after
Before and after
Before
No way to match personal positions to candidates across a whole ballot
Cross-platform app on iOS, Android, and web
No tool compared candidate stances across local, state, and federal races
Short issue quiz surfaces aligned candidates in every race
Down-ballot races — school board, city council, judges — went largely uninformed
Down-ballot races covered as deeply as top-of-ticket
Civic tools clunky, outdated, and desktop-only
AWS infrastructure holds through election-season traffic spikes
Hours of independent research per candidate
Voters arrive at the polls informed
Common questions
Common questions
More case studies
Get started
Ready to see what AI
can do for your business?
Book a free 60-minute consultation. No sales pitch — just honest answers about where AI fits.