Housing.com Takes A Leap With Natural Language Processing-Based Search For Users

Leading real estate portal, Housing.com has used analytics to heighten the level of quality it rolls out to its users. The portal has delved deeper into Natural Language Processing (NLP) to ascertain what prospective homebuyers and tenants are looking for and to deliver the best search experience. In a short span of time, 70% of users have been using NLP and it has been an instant hit.

Communication is a great tool but there could be as many ways to communicate as there are people. It becomes important for businesses to understand the pulse of their consumers and at Housing.com, the analytical approach has helped release the latest feature which is the NLP-based search engine that lets users type in their requirements in the natural language and still get relevant results. Released in August 2018, Housing.com has achieved the distinction of being the only real estate portal in India to do so. Consumers can now expect a Google-like search experience when they search for a property on Housing.com.

Better search, higher conversions

The research that went behind this feature was backed by some interesting findings. While most real estate portals are helping users arrive at the most relevant property through filters, Housing.com gauged that more the filters, the higher were the conversions. But, users would definitely be looking for a quicker way to arrive at the desired property.

Based on this finding, Housing.com now allows users to search in their natural language. For example, if one is looking for a “2BHK in ABC locality between 19 lakhs to 22 lakh”, they can just type this entire instruction and the search will lead to the relevant results without selecting the filter. The feature also helps users with auto-suggestions based on budget, configuration and amenities.

Additionally, NLP has helped the business with 2x the number of filter usage by users than before as well as a significant jump in conversions.

NLP has been designed for not just homebuyers but even those looking to rent.

Behind the scenes

The idea behind introducing this feature stemmed from Housing’s participation in the Global Hackathon “Code Jam 2018” with Moondragon – the NLP-based search engine and it became Housing’s winning entry at the global hackathon.

The brain behind Moondragon realised that most internet users were now autowired to search in their natural language and therefore the need to give them what they are used to is important. Real estate solution providers had never before speculated this need but analytics showed that in decoding the information entered by users lies the strength of the organisation. The company decided to give a type-ahead experience to all its users and the analytics that went behind it suggested that users either start their search with location or with configuration. The NLP model provided the position of user’s identified filter and helped re-order auto-suggestions for a far smoother user experience. The traditional way of searching shepherded a number of drop-offs at each step. NLP settled the difficulty by providing all the filters reciprocally.

Analytics and technology have led to better predictive capabilities and it is a win-win for our users and the business as well. However, we have never stopped challenging ourselves and this is exactly why we are now the only real estate service provider with this distinction.