Here’s the pitch deck used to raise a $4.4 million seed round for an AI chatbot looking to transform how people find apartments
This story is available exclusively to Business Insider subscribers. Become an Insider and start reading now.
- RealFriend, launched in 2018, is a startup that created a chatbot to simplify the apartment-hunting process.
- RealFriend first went live in Tel Aviv. After a year-long beta test, it is now officially launching in New York.
- RealFriend’s cofounders shared the pitch deck that helped them raise a $4.4 million seed round earlier this year with Business Insider.
- Visit Business Insider’s homepage for more stories.
In the hottest markets, finding the right apartment can feel like a full-time job, wading through fake listings and apartments that don’t fit the needs of the renter. Even a rental broker, who can help with some of this, has a smaller, siloed view of the market because they’re tied to a specific brokerage and a limited amount of listings.
That problem is what convinced long-time collaborators Omri Klinger and Hadar Landau to apply the AI-chatbot they had been building to the world of real-estate rentals in 2017 with the launch of RealFriend.
RealFriend first went live in 2018 in Tel Aviv. In 2019, the chatbot spoke to almost 54,000 renters in a city where 50,000 apartments are rented a year. The company also launched a beta run in New York last year, renaming the chatbot from Dooron, its Israeli name, to Luke, as it began to operate stateside.
RealFriend raised $4.4 million in a seed round this March, with backing from Gaia, a NYC real-estate investment fund; Eyal Waldman, CEO and founder of Israeli-American computer hardware company Mellanox Technologies; Israeli venture capital firm F2; angel investor and former SVP of product at WeWork Ron Gura; and German media company Axel Springer.
Now, RealFriend is launching more widely in New York, with the hope of helping renters navigate the famously challenging New York rental market as it rapidly changes due to the coronavirus.
Business Insider spoke to Klinger, the company’s CTO; and Landau, the company’s CEO about their pitch deck and why they think chatbots will massively change real estate.
Here’s the pitch deck RealFriend used to win over investors and raise its seed round.
(Disclosure: Axel Springer is Business Insider’s parent company.)
Realfriend first launched in Tel Aviv, and has now been running a beta for over a year in New York.
The company gave the chatbot a traditional name in order to facilitate more natural conversations with the bot.
“You tell Luke things you tell only to friends,” Landau said. “In real estate, which is an emotional process, having someone who has your back is what you’re looking for.”
Landau and Klinger have been working together for 15 years, and first met in the Israeli Defense Force.
Israel is one of the world’s leaders in successful startups, with Tel Aviv having the most startups per capita outside of Silicon Valley.
Startup experts told Business Insider last year that the IDF is one of the main reasons for this, because of its use of cutting-edge technology and the discipline that is instilled in its soldiers.
Landau and Klinger met in the IDF in 2005, and by the time they launched RealFriend, they had already launched one startup together, OneFeed.
OneFeed was a browser extension for Chrome that turned the “new tab” page into a dashboard with personalized news, email notifications, and social-media alerts.
Early experiments with chatbots and a bad experience apartment hunting led the two to make real estate their focus.
After selling their previous startup in 2015, Landau and Klinger began to experiment with chatbots. It took them more than a year to create a conversation engine for the chatbot that satisfied their demands, which coincidentally lined up to when Landau was searching for a new apartment in Tel Aviv.
He spent hours sorting through many listings, some of which were fake, to find potential apartments to rent. Afterwards, Landau realized that apartment rental would be a perfect application for the conversation engine they had built.
“We realized that what we built could be the future of real estate,” Landau said.
RealFriend’s central concern is making sure its chatbot can interact with people as if it were actually someone they could speak to.
The company used the above chat log, which shows a user “introducing” Luke to their mom, as an example of the bot’s conversational skill, especially in relation to other rental search options online.
The company used chat because it is more personal than other options, is accessible to more people, and allows for more flexibility in the amenities and requests it can accommodate.
Key to their choice to use a chatbot was the more personal nature it brings to the real estate rental process, which is one of the most important financial and personal decisions that people make.
Chatbots are also more widely accessible to users who aren’t as comfortable with tech and don’t want to learn a whole new app. Landau gave the example of an older customer who uses voice-to-text to communicate with Luke, something that would be significantly more challenging if Luke were an app.
Chatbots are also more flexible, allowing the system’s database to easily add other considerations or features that might be hard or impossible to find on a traditional website, and would be a user-experience nightmare to code into a traditional website.
“You can tell it things that are not in an app, like an open kitchen layout and south-facing view from my living room,” Landau said. “This is something you wouldn’t be able to search on StreetEasy.”
This sample chat shows Luke’s ability to suggest properties on the market, provide data about the market, schedule showings, and iterate on feedback.
The chat begins with a user telling Luke what their requests are for an apartment, leading Luke to provide a potential option. Luke’s engine, which combines information from many apartment listing sites, is able to give data about its affordability compared to the market as well.
The app then schedules a rental showing, and then after the showing, asks for feedback. The user explains why they want to pass on it with quite specific details, prompting Luke to suggest another apartment that matches one of the user’s concerns. Those concerns are also logged into RealFriend’s backend database, so that the company doesn’t recommend this apartment to people with the same concerns.
This slide helps to explain why RealFriend is using a chatbot, and why it sees such an opportunity in the US.
Despite having no marketing beyond word of mouth, the company became quite popular in Tel Aviv.
Dooron spoke to more people in 2019 than there are apartments being rented in Tel Aviv.
The company’s New York beta, launched in 2019, now has 2,500 monthly active users.
The company soft-launched Luke in the notoriously-challenging New York rental market in March of 2019, and has seen a 25% month-over-month growth in engagement.
New York was always planned as another market to launch in, as Klinger was actually living in the city when the pair decided to apply their conversation engine to rental real estate.
Luke and Dooren’s Net Promoter Scores are higher than those of other, larger AI assistants and dwarf other real-estate portals.
The company’s service is free, but they’re hoping to eventually make money by charging agents a referral fee of 25% of their commission.
As of now, Luke and Dooron are totally free, but the company plans to eventually charge referral fees from brokers.
“We collaborate with brokers, and we do some of the tedious work they used to do to save them time and help them focus,” Landau said.
While the company looks to eventually make money through referral fees, not too dissimilar from models like Zillow’s search portal, they don’t plan to actually sell their services to a whole brokerage.
Landau said that they’ve received a lot of offers to partner with brokerages, but have decided not to because they want Luke to be a neutral arbiter who helps out the person looking for an apartment and is not financially aligned with a specific broker or apartment.
“Staying unbiased is important,” Klinger said.