The web has fundamentally moved us forward. Most notably, it has given us access to the world’s information and there are currently three main ways we discover an ever-growing amount of content: news, social networks, and search. We believe that a fourth category is missing: relevance—a calm, quiet place where we take a step back to focus on what’s most relevant to us.
News delivers what’s happening in the world right now. Social networks let us know what’s happening with our friends. Search is great at finding the needle in the haystack.
But how do we discover things from around the web that are new and relevant to us? There’s a plethora of curated content—from niche sites over Twitter to newsletters. While it’s interesting to see the web through the lens of the curator, our interests are unique and relying on a few curators alone is likely not diverse enough. If we add more sources to our mix (for example, following more people on Twitter, or using an RSS reader), we increase diversity but also noise.
We need a better filter.
We read the news to see what’s happening in the world right now. We check our social network to see what’s happening among our friends. We search when we know what we’re looking for. And to see what has happened around the web that’s most relevant to us, we go to our relevance place.
Let’s first define what we want. We’d like to get a list of the most relevant links: more on our biggest interests, fewer on others. We want to see what people we follow find relevant. The list should be novel and accurate, but with a sprinkle of serendipity. It should burst the filter bubble. But it should also be limited.
This can be solved algorithmically, but a new metric is needed. A metric that takes different signals into account, and evaluates things differently. Instead of clicks, we leverage signals that better capture relevance. Instead of age in seconds, we favor long shelf-lives.
The solution to the second problem—too much to handle—is part human, part computer.
The human part is a simple triage: do we want to save a link—fire-and-forget style—to keep and find again when we really need it, or do we actually want to read it? This is important, as it reduces our reading lists significantly.
But it only works if the system re-surfaces links when we actually need them. That’s the computer part. It’s a lot about the right link at the right time. But it’s also about the right form: when we commute, we might prefer audio; when we’re in a hurry, we might prefer summaries.
We’re building Refind to tackle the relevance problem. Refind is a community of movers and shakers who discover, save, and read what’s worth their attention.
Every day we analyze an ever-growing amount of new links and create a magazine of the most relevant ones for you.
For every new link, we compute your personal relevance score and show you up to 50 that make the cut.
Save great links from around the web with one click—with our browser extension on your computer or our share extension on your phone.
Search the web as you always do. Refind highlights links you or your friends saved when you search on Google.
Our smart reading list helps you read more of what’s relevant to you. Links you don’t read within two weeks will move to Someday, sorted by relevance. Soon is always small and manageable, and if you ever find time to catch up, you’ll find more on Someday.
On iOS, you can access your reading list anytime, anywhere—even when you’re offline.
Also, we’re trying to bring you the content in the right form: when you commute, you can listen to the story; when you’re in a hurry, you can read a summary.
Refind is a work in progress. It’s used by thousands of engaged users and we continuously improve Refind based on feedback. We’re grateful for all the input we get from engaged users—people like Chris Messina have had a tremendous impact. We’re 1% there and have a long way to go.
Two years ago, we set out to solve our own problems—handling the volume and cutting through the noise—by writing code. But along our journey, we continuously stumble upon people that shape what we build. It’s a long list and we’ll portray them on our blog at some point. Here are three.
We’re passionate about the web, search, data science, and so many other topics. Fundamentally, we’re fascinated about what can be done with an idea and a computer. We want to move the edge. There’s so much to be done, but we focus on one task at a time. Let’s get to it.
We’re hiring! Reach out if you want to contribute to our mission. At this point, we’re mainly looking for mobile developers, data scientists, search and scalability engineers.