Machine learning is a subset of artificial intelligence.

Machine-learning will disrupt online classifieds the same way the internet did two decades ago to print classifieds, REA Group (ASX: REA), Frontier Digital Ventures (ASX: FDV) and Google (Nasdaq: GOOGL) told attendees of Australia’s first Wild Digital conference recently.

Machine-learning is the branch of artificial intelligence (AI) concerned with computers learning to act in ways without being explicitly programmed to do so, and it’s going to transform how consumers search for property, cars and jobs.

In the very near future, instead of punching the details of a car you’d like to buy into a classified site on your phone, you’ll be able to tell your smartphone you’re looking for a car, and it’ll already know exactly the kind of car you’re looking for. Soon enough, you won’t even have to tell it you’re looking for a car, because it’ll already know that too.

Google called this personalization, and according to Google’s Australian managing director Jason Pellegrino, this will be the next battleground for online businesses (not only classifieds).

“Companies, services and products will live and die in their ability to deliver personalized services, because delivering the same services to all consumers all at once, or even localizing them, will not be good enough.

“And the only way to deliver that personalized service, is to use all the tools that are available through machine-learning.”

Already 20 percent of Google searches in the U.S. are voice searches, which Pellegrino said has grown from “very insignificant amounts” over the last 18 months.

The uptick in voice search is due to the machine-learning algorithms sitting behind the Google search app, which makes it able to learn and understand natural language and to understand context.

Pellegrino said machine-learning requires three things:

  1. The ability to compute really, really quickly;
  2. Huge data sets about products, services and consumers; and 
  3. Constant connectivity. 

Online classified businesses are at a significant advantage compared with other digital companies, in the sense that they already possess massive amounts of data, be it about cars, property or employers and employees.

And, in the frontier regions of Southeast Asia, which Frontier Digital Ventures (FDV) founder and CEO Shaun Di Gregorio said is a mobile-first economy, because very few people use desktop computers or laptops, there’s great potential for classified businesses in those markets too.

Shaun Di Gregorio

“In a lot of our countries, traffic to the mobile site is 90 to 95 percent of total traffic. And it’s driven by really cheap smartphones,” Di Gregorio said.

There’s also very little governance and consumer protection in these regions, so classified companies have also come to facilitate transactions between buyer and seller, further adding to the wealth of data amassed by these online portals.

“These guys end up becoming the go-to platform, not just to look at houses or cars, but as a trusted intermediary. So they get to do more than just sell ads,” Di Gregorio said.

“They’re not becoming property agents, they’re not becoming car dealers, but they follow the buyer right to the point of working with the seller.”

According to Pellegrino, over 350 product launches from Google in 2016 had machine-learning at the heart of those products.

However, internet speeds still remain the biggest barrier to machine-learning.

REA Group, which is integrating machine-learning and robotics into its platforms, sees Telstra’s (ASX: TLS) work on the 5-G network in Australia as being crucial to its success.

Telstra estimates 5-G will arrive in Australia around 2020, but the company plans to roll out a trial 5-G network at the 2018 Commonwealth Games on the Gold Coast, so it could be earlier.

When that happens, REA Group said, there’s going to be an explosion of rich media utilizing machine-learning, and they want to be poised to be able to meet that.

So watch this space.

The Wild Digital conference was held at the Sofitel Wentworth in Sydney last month.

Print Friendly, PDF & Email

Related Articles