Property Website Tells Home Buyers Where to Find Property

Share Article

Gravity (http://www.g.co.za), South Africa's property search engine has found a way of predicting where its visitors will most likely search for property.

We'd grown to about 100 000 listings spread over more than 3000 suburbs and we wanted to find a faster, more user-friendly way to get our visitors to the properties that would be of interest to them

Grant Leigh, CEO of Gravity Communications, believes his company has found a unique solution to a problem many large property websites face: "We'd grown to about 100 000 listings spread over more than 3000 suburbs and we wanted to find a faster, more user-friendly way to get our visitors to the properties that would be of interest to them".

The normal Google-based street-level-map search approach adopted in the US and UK is not a viable option in South Africa because agents don't include street addresses in their listings and also because Google Maps isn't as granular in South Africa as it is in those countries. There are also other problems with the map-based approach.

Home buyers are often not interested in a suburb purely because of the area it’s in - they may be looking at suburbs in completely different parts of town or, worse still, they may be looking at different towns across a wide geographic region.

Location or geo-based solutions also don't take into account the type or price bracket of suburbs. For example, in Cape Town, the residential suburb of Milnerton borders on Montague Gardens - a commercial area. A location-based model would present listings in both locations to the user, but Montague Gardens listings are nowhere to be found on a Gravity property search in Milnerton. In fact, the Gravity model predicts that buyers are more likely to look for houses for sale in Blouberg Rise or even Big Bay over 10 kilometres away!

"The problem just didn't seem solvable, so we elected to start work on a location-based system. Then, one day as I was explaining to a colleague the extent of the mammoth task ahead of her, it struck me. Via the search data they left, our hundreds of thousands of visitors had been telling us exactly where they wanted to live since the day http://www.g.co.za was launched."

After sifting through the data and coming up with an algorithm that best harnessed all that collective intelligence, the first release was quietly launched several months ago. Visitors could enter a single suburb of interest and the Gravity site would suggest 20 further suburbs they may wish to explore. Encouraged by the results of this release, Gravity have now brought out a more advanced version. On the Gravity homepage, the user is able to enter multiple suburbs (in the "Quick Search" box; comma-separated). The results of this initial search are displayed to the user as well as a list of the most relevant suburbs the Gravity algorithm predicts will be of interest to the user.

Have any other websites come up with this solution as well? "I'm not sure; not as far as I can tell though. The idea may be new, but the concept of User Generated Results is old in the search world. I find it ironic that although this may seem like a fairly technical solution, it’s based almost entirely off human intelligence. I think this is a direction in which internet search as a whole is moving."

About Gravity:
Gravity (http://www.g.co.za)] is a South African Property Search Engine that is free to estate agents and homebuyers alike. It automatically scans estate agent sites, gathering their listings and placing them in a single, central database for users to search.

Contact:
Grant Leigh (CEO)
Gravity Communications
0824021031
grant(at)g.co.za
http://www.g.co.za

###

Share article on social media or email:

View article via:

Pdf Print

Contact Author

Grant Leigh
Visit website