UPDATE — We just launched our iOS app but thought this older post would provide some insight!
Over the past year, I’ve been working on a project called dorsia.io, a website that aggregates recommendations from food critics, travel writers, and established bloggers into an easily searchable map.
Dorsia’s was formed around the idea that there is already too much content in the world. When we ask Google “find me a restaurant in SoHo”, we don’t want an answer derived from the the collective intelligence of the internet. We want an answer from a subset of people who think and write about restaurants all day.
Why Did I Build This?
Dorsia was built out of frustration over spending countless hours researching things to do while on vacation. I always found it difficult to get a grasp of what’s going on in a place I’ve never been to if there wasn't a recent NY Times 36 Hours piece or a No Reservations episode on file.
When I travel, or just explore a new neighborhood, I’m looking for a narrative, something that jumpstarts my exploration. For example: Jackson Heights, Queens: Diverse and Evolving. The title alludes to an adventure off the beaten path, and even though this is an article in the real estate section of the NY Times, it brings context to a part of NYC that is often overlooked by tourists and residents alike.
I wanted to build something that helped me discover these narratives while using an interface that felt familiar.
The Current Way of Doing Things
In my experience, there are really only four ways to figure out what to do in a given location:
- Ask for recommendations from friends.
- Read articles about the place.
- Ask the internet.
- Wing it.
I’m a big fan of options one and four. Most of my friends travel a lot and and have similar tastes to mine, so it usually works out. Winging it, while exhilarating, is not for everyone. It’s led to some cool discoveries, but in other cases has left me wandering aimlessly for hours through lifeless residential areas.
Option three is the worst, because you end up having to sort though shit like this:
If you’re willing to put in the work, option two has the best ROI, but it is also flawed. After thinking about it a little more, I came to a few conclusions about travel research:
- Research is hard because there is too much content to sift through. It’s difficult to figure out what’s happening in a city when articles about new places are buried under a sea of search-engine-optimized posts from a few years ago.
- The clout of iconic sights tends to create a digital fog that hides under-the radar locations that are popular with locals.
- Ranking places is worthless because it either relies on averaging the sum of user-generated reviews or is one person’s opinion.
- Among restaurant and bar critics, there is a general consensus about what’s worth visiting. Their recommendations will almost always be better than a “best hits” list from a social app.
These learnings helped me determine what I wanted in a travel discovery tool. What I was seeking was a way to easily search well-written articles, to parse through a big list of things to do and pick what interested me. I wanted a way to filter out the noise and avoid tourist traps. A virtual “fixer” that could point out all the cool shit in town.
Scratching My Own Itch
After googling things like “Rotten Tomatoes for travel” and coming up shorthanded, I decided to start designing a site that would solve my research problem.
As a programmer, my first attempt at building this was to scrape the internet for reviews from a list of vetted sources. Google actually does a pretty good job at this. They aggregate reviews for restaurants from major media outlets. The problem they aren’t solving is discovery. Google Trips relies on user-generated ratings to push the “best” restaurants and sights up to the top of the list. Travel discovery doesn’t work well when designed as an an algorithm, because computers don’t have taste.
When an algorithm plans an itinerary for a trip, it barfs out something like the picture below; Google’s “If you only have a day in [NY]” list. As you can see, it’s an itinerary that only Spock would enjoy.
I realized that the challenge in building a great travel discovery site was not simply an engineering problem. It required a hybrid editorial/engineering approach to solve.
For the editorial side of things, we ended up hiring people — researchers, if you will — who have ties to the local bar and restaurant scene in the cities that we cover. Bartenders, servers, cooks, and bloggers usually have amazing recommendations.
Our researchers add locations to Dorsia based off of what their friends and friends-of-friends are recommending. We then ask the researchers to find some compelling writing to back these recommendations up. Our theory is that if two or three well-written articles are praising a new restaurant, and one of our researchers recommended it, it’s probably worth your time to check it out.
That’s why we cover everything from Michelin-starred restaurants to dive bars. Both types of places are great, and sometimes it takes an expert to tell you why.
I want Dorsia to be an egalitarian concierge service, a way to help people filter out the noise when planning their outings, a source of information that consistently points them in the right direction.
The first iteration of the site is focused on restaurant and bar recommendations. Soon we will be adding shops, cafes, sights, and events. We also have plans to expand into other cities.
Thanks for checking Dorsia out! If you have any feedback, or just want to say “hi”, you can reach me at andrew[at]humidresearch[dot]com.