Inspiration
Duckvan started when Sanford and Tony, who are currently roommates, had to find a new apartment. We searched through multiple websites - Trulia, Zillow, Craigslist, and more. Once we found an apartment we liked, we would cross reference Yelp and other review sites to make sure we wouldn’t fall into a trap. Finally, we would visit a the top apartments out of the hundreds we considered. The visit appointment process involved many back and forth phone calls with property managers. Overall, our apartment hunt took a solid week of effort. We would much rather pay someone to do all this work for us -- such as how rental brokers work in NYC. However, these brokers charge up to 2 months rent (!!!). Duckvan makes it affordable to outsource your apartment hunt without shelling out big bucks.
What it does
Duckvan will hire “agents” who will use our technology to help apartment seekers find their next home. These agents will NOT be real estate professionals, but rather ordinary people who want to earn some extra money. Our technology will give them apartment hunting superpowers that will let them help users find a place to live!
On the consumer side, a user downloads the mobile app and fills out some basic info (#beds, #baths, location) as well as some personal preferences (e.g. tradeoffs between price vs. quality vs. location). The user’s request will be posted to the Duckvan system where agents will use a website to choose jobs they want to start. The agent have a quick chat with the user to clarify any questions about the search. The agent will use listing sites and review sites to curate a set of apartments that match the user’s requirements and personal preferences. Once they’ve found the best apartments, they will simply paste a link to each apartment listing and write some tags that indicate how that apartment is relevant to the user. The user will get a notification on their phone and the app will show the user the curated list of apartments.
The user can directly book an appointment with the property owner for rooms they like and delete others from the list that don’t fit their needs. The user can also chat with the agent to explain why some rooms were not a good match -- the agent can then take this feedback into account and refine the list.
How we built it
We created an iOS app for users to submit requests and chat with an agent. We also created a web application using NodeJS for agents to see list of requests, sends user apartments and chat with user. Our web scraper system was built using Python + JavaScript.
Challenges we ran into
Getting data sources into our system is hard. We developed a template based web crawler that figures out relevant information by looking at the structure of web pages on apartment hunting websites. The agents can simply paste a link to any listings page (Craigslist, Trulia, Apartments.com, etc) and the scraper will intelligently switch between a set of templates for each site. We learned that collecting the entire breadth of apartment data is an exciting challenge, it is a combination of fine tuned data analysis and human expertise. We’ve also found out that the most crucial data about apartments have to be elicited from apartment management. This creates an interesting value proposition where we can inform apartment management what kind of data customers are interested in and in return we gain crucial information that didn’t exist before.
Accomplishments that we're proud of
Getting the end to end flow working in 48 hours :)
We showed screenshots of our app to friends in SF and NYC, almost all of them wished it was live in the App Store!
What we learned
Getting feedback and iterating upon it is very important. So we stood outside Starbucks and asked customers in the age range 20-30 years old for feedback on our app. Most potential users loved the fluid design and thought our app solved a very important problem. However, some users complained about the slow chat interface, so we changed from using Sinch’s chat API to SendBird’s chat API.
What's next for Duckvan
We need to figure out the unit economics and see if we can find a pricing model that works well for both agents and users.
Become the one stop shop for anything apartment related that requires human interactions. Need to setup electricity, internet, and other utilities in your new room? Duckvan it. Need to call maintenance to repair your fridge? Duckvan it. Remove all friction in the process of finding, renting, and moving between apartments.
Log in or sign up for Devpost to join the conversation.