Outdone

In 2020 we sourced the help of a data scientist freelancer and began working with him to build the gift recommendation engine using the data we gathered. Well, a couple months into the project, he went silent. Ghosted us. (Weeks later he ended up apologizing, but the damage was done.) And we had already gone back to the drawing board.

By the time we found his replacement, a data science team based in Utah, it was spring of 2021. Luckily, this team proved to be incredibly responsive and perfectly skilled for the task at hand. The site and MVP recommendation engine took four months to build and wrapped up right on schedule. We wanted to give ourselves a bit of lead time ahead of the peak holiday shopping season. So here we are — eager to test the concept and bring Outdone to the world!

With a few family birthdays and the holidays looming, my co-founder (and roommate at the time) and I were stressing over what to buy our siblings and parents. After a couple days of gift deliberations, it dawned on us that there had to be a smarter way.

So one night we brainstormed how the problem could be solved. Being a researcher by day, my co-founder, Hugh's first instinct was to build a consumer survey to understand what people love and why they love it. My full-time job is with a martech firm that leverages machine learning, so I knew if we could build a strong enough dataset, we could use ML to build a powerful recommendation engine. And with that, the Outdone journey began.