SwayMe started in 2017. Two Columbia engineering alumni met at a reunion dinner, incorporated that August, and built a version of this: two photos, a question, a vote. It launched. It didn't go anywhere, and by early 2020 we'd shelved it. The problem wasn't the execution — it was that two-photo polling is a feature, not a product. There was no reason to come back tomorrow.
The idea wouldn't leave us alone, though, because the need is real. People make a dozen visual decisions a week and the options are all bad: text five friends one at a time and wait, throw it on your story where it vanishes in a day, or ask strangers on Reddit who don't know you.
In 2025 we started over. The thing the first version was missing turned out to be groups — the ability to ask a specific set of people rather than everyone or no one. That changed what the product is. Around it we built an endless feed worth scrolling, streaks and challenges that make voting a habit, and the beginnings of the AI layer that routes polls to people who'll have an opinion worth having.
We're a two-person company in New York, early and deliberately so. Right now SwayMe is free and we're focused entirely on getting it into the hands of people who'll use it daily — students first. Longer term we think the most valuable thing here is the data: what people actually choose, rather than what they tell a survey they'd choose.