The real estate market has long since been hailed as being a natural fit for the application of artificial intelligence and machine learning models. This is due to its fragmented nature, filled with brokers and intermediaries looking to get in the way of a tenant looking to buy or rent a house. More specifically, in India, broker culture is so widely enforced to a point where brokerage has become accepted as a social norm.
Looking to change this, Saurabh Garg, Amit Agarwal and Akhil Gupta founded NoBroker.in in early 2014. As the name suggests, the site was aimed at removing brokers from the real estate market as they do not add value to the real estate market. Instead, brokers function on selfish means and do not give the customer a good match for their house search.
NoBroker entered the market with an aim to introduce an AI/ML solution to the widespread broker problem in India. Today, it is one of the leading data-driven real estate companies in India, with over 1 lakh properties being posted over the past month. The website has also introduced various AI-based products such as Rent-o-meter, livability score and transit score to enable a democratic house search for consumers.
To find out how NoBroker evolved to be the first and only AI/ML driven real estate company in India, Analytics India Magazine reached out to Akhil Gupta, the Founder and CTO of the company.
The Germination Of A Broker-Free World
In the early 2000s, the founders saw that services across verticals were beginning to move into hosting a website. While these platforms had the branding of the companies offering them, they were just a website who connected users with brokers with no real use of data. This drove them to create NoBroker, to eliminate the broker altogether from the real-estate process, and adopt a data-driven approach to deliver useful information to the consumer.
Gupta mentioned how NoBroker facilitates all transactions on the platform, allowing them to gain a huge number of leads in the process. According to him, this will help build the machine learning and AI which will be able to give insights to the users. Demonstrating the data-first approach of the company, Akhil stated, “One thing we made sure of is that we have all kinds of data, we don’t lose out on any data. So we designed our system that way, we had a huge amount of data in the system.”
The company began operations in Mumbai, later expanding to multiple cities such as Bangalore, Pune and Chennai. Currently, the company is headquartered in Bangalore.
Data-Driven Features To Find The Right House
The point of NoBroker was to create a platform where brokers would not be allowed, but this was only the beginning of AI being used in the service. Reportedly, NoBroker found problems with the owner not picking up the tenants calls. Noticing this, NoBroker quickly initiated a change that would fix this issue faced by the consumers on the app. Akhil said, “We introduced something called “Call Alerts”. Now what happens is, the owner has the app, and the user has the app, and when the user calls, even when the owner doesn’t have the user’s number, there is a card that flashes up telling the owner that a user from NoBroker is calling them.”
Reportedly, this has increased call pickup rates and the number of deals being closed. Other AI driven solutions also exist at NoBroker, such as the Rent-o-meter. The Rent-o-meter can accurately tell how much the rent for a property should be, given about 70 attributes of the building. They are also creating a model that will predict the selling price of a property.
The Rent-o-meter functions on the principle that “every property is unique, and no two are the same, even if they are in the same building”, said Gupta. He elaborated “What we have done is we have around 70 attributes of the property and we have created a prediction algorithm. So what it does is, at a street-level accuracy, it predicts the rent for a property. This is a self-learning ML algorithm which will get more and more powerful as more transactions continue to happen on the platform.”
NoBroker also has other AI-driven solutions, such as a transit score, a livability score and travel times. For example, the transit score is an ML-based scoring system that looks at factors such as nearby bus stations, metros and waiting times for cab hailing services. The livability score takes into consideration the number of nearby hospitals, amenities, supermarkets, malls, cinema halls and other entertainment based services.
Analytics In The Big Data Revolution
Speaking of the future, Gupta said, “We have to build solutions where we can tell a developer, your next project should be in this locality because we see high demand and very less supply available in that area.” This itself demonstrates the data-driven approach NoBroker has towards solving for the real estate vertical. However, it seems that they want to continue to find newer solutions to increase the value of NoBroker to their consumers.
Agarwal echoed a similar sentiment who believes that analytics is constantly evolving as a field. Regarding this, he stated, “With data, it gives you a lot of visibility to solve the problem, but it opens up new problems.” One of these problems is also taking responsibility for the data collected by the company. For example, Akhil stated that NoBroker does not track location data of the users, because they simply don’t need it.
However, owing to the nature of the platform, NoBroker’s data is accurate. This is due to the fact that there are no brokers on the platform, and all data is provided only by the owners and the tenants. NoBroker also functions on collecting data from the real estate transactions that take place on the platform. “When this happens we are emitting such unique data. Now if the customer comes and searches for the property. I know what his demand is, where is supply available. This is an extremely valuable proprietary data what we have. What we have done is make use of this transaction data to power our Rent-o-meter,” Gupta said.
However, Gupta echoed the spirit in which the company was founded when he said, “What we do and what we want to do is use ML to remove all the information which is useless, bring in a transparent democratic way for the customer to choose the house and for the owner to find a tenant or a buyer for a house.”