TECHNOLOGY SPECTATOR: Smarter real estate bets

Strong headwinds facing the Australian property sector have forced real estate groups to get smarter, with many turning to technology to prise value from the market.

Technology Spectator

The latest housing data from the Australian Bureau of Statistics have highlighted how global economic headwinds continue to bite the Australian property sector. According to the ABS, housing finance in July fell 1.8 per cent in seasonally adjusted terms to $20.05 billion –- down one per cent on the same time a year ago – and well down on the peak at the start of the year. And while clearance rates in Melbourne and Sydney hover at 60 per cent, an increase in properties coming on to the market in the traditionally busy spring season will dampen prices.

The flat conditions mean property groups must be smarter about where they place their bets and some are implementing the best tech solutions at hand. Analytics is playing a bigger role in finding properties that are not appropriately priced, seeing business opportunities in new geographies and providing a value-added service to nearly 1000 franchised operators in the Ray White property group, says its marketing senior analyst, Nathan Krisanski.

He says the MapInfo product it used prior to switching to business analytics provider Tableau lagged the group's direction.

"It cost a lot of money to maintain and was difficult to update," Krisanski says. "We were looking for something a little more flexible to deliver."

That was five years ago when a key report was market share by suburb, "pie charts on top of a map", he says.

According to Krisanski, the decision to implement Tableau "led to great discussion with business owners and corporate about office performance and opportunities for new offices". And with reports that last year 10,000 agents left the industry – there were 72,000 in the 2006 census and 50,000 last year - such services offer a big property group an enticing carrot to lure strapped talent to its fold.

Ray White provides back-office services to franchises and charges out value-added services such as websites, client-relationship management and hosted productivity and collaboration tools in Google Apps, which it signed on to at the beginning of the year.

The property group has in its repository up to 22 million rows of data by 12 million fields – "if viewed as a big Excel table" – in SQL database tables from which realtors pull customised reports on demand over the web. They are used for internal pricing and distribution to prospects and typically include suburb demographics, recent sales with photos and descriptions of houses and statistical information to inform realtors, buyers and sellers. Agents log on to a webserver to request the custom reports delivered as branded PDFs, which are typically generated and emailed back a few minutes later.

Unusually, Ray White hosts its own Tableau server inhouse (it's often hosted or in the cloud) and the database on two virtualised servers, although it is considering cloud services.

"We have a fully virtualised data warehouse, and Tableau runs on a four-core machine with 8GB of RAM as a single instance; the Tableau guys have told us its the lowest-specced machine they've seen it run on." The database runs on a two-core machine with 8GB RAM. So,the memory footprint isn't that great.

By the numbers

Ray White's largest dataset has eight dimensions including client and property type, Krisanski says.

Head office pulls in data each night from internal systems such as CRM and sales reported by agents and monthly updates from property price reporting services RP Data, Australian Property Monitors and Price Finder.

"We have those three sources we compare against to understand the completeness of the data and if there are holes" in it, he says. "When it comes to different states and suburbs, they vary wildly in their relevance."

The past quarter's data is critical to understanding the market and this is where the sales data providers compete because after that state governments provide empirical data, putting each of the private providers on the same level.

Krisanski identifies a problem that besets Australian residential property sales in how data is collected and reported. In the US, services such as Zillow provide an instant, god's-eye view of the market at a click or touch but a deficit of Australian standards and laws demanding prices be reported accurately and promptly by vendors and their agents on a non-discriminatory basis makes such a service impossible here.

"In the US, they have the MLS (multiple listing service) which is an open, online information system for all property sales in one behemoth of data," Krisanski says. Although there is no authoritative repository for MLS data, the use of the Real Estate Transaction Standard protocol, extensible markup language (XML) and sharing conventions between agencies and representative bodies provides a framework for data reticulation to interested parties.

"They have an industry body that collects that information and feeds it back out - a very different ballgame to back here where there are very big costs involved collecting that data and no standardisation of how it’s manipulated and presented back,” Krisanski adds.

"For Australia, there isn't an open source for that information and, without having expensive arrangements with RP Data or Price Finder, you can't do that."

For instance, the Real Estate Institute of Victoria levies royalties of up to $18,000 a month to buy sales information, "quite a significant cost", he says.

The final aspect of solutions like Tableau is the inherent simplicity when it comes to installation and operation. Tableau arrives with a Windows installer file that runs its own Apache webserver and Postgres database.

"Double click on the installer, and click-click-click, you're done", Krisanski says.

With active directory for authentication and management integrated into a purpose-built administration console, groups like Ray White have found a whole new way to look at the property market.

As Krisanki points out it’s not just about getting the most out of the available data but also presenting the material to agents in an appealing way.

"It's the same sales information from RP Data but presented in a way that no one else can with Ray White branding."