The term ‘big data’ is often misperceived, according to Airtasker’s Adrian Letchford.
The “big data” buzzword may have slipped out of fashion but the idea of distilling meaning out of vast volumes of consumer data is more relevant than ever, according to industry experts speaking at the Future Examined event, hosted by The Australian and Deakin Business School.
Sydney-based task marketplace Airtasker’s senior data scientist Adrian Letchford said dispelling common misconceptions about the term should be the first port of call for any organisation looking to dive into the world of data analytics. “One of the things I think about when it comes to big data is it’s not just big, it’s not just huge volumes of data.”
According to Dr Letchford, the core focus of any strategy needs to be around outcome and getting the big picture.
“One of the big differences that we get with big data is that we can see the whole picture. We’re not just looking at a small sample, we’re looking at the entire picture, the entire playing field,” he said.
This is especially relevant in the current environment, where the efficacy of traditional data analysis, reliant on a small cross-section of the overall population, has been brought into question.
“Say a polling company right before an election goes out and polls, say, 2000 people, and then they collect all these opinions together, which is then reported in the news as who we think Australia is going to vote for.
“What we all know is that they often get this wrong, and they get it wrong because the level of certainty behind that analysis is really, really low,” Dr Letchford said.
However, increasing that degree of certainty isn’t just about scale, it’s also about putting ideas to test as quickly as possible and centralising it.
“Get all your data together — it makes the world of difference”
That’s what Airtasker is doing through its recent investment in A/B testing, where two versions of an idea can be compared to determine which one works better.
“This is something that has been really exciting for us because big data isn’t necessarily about huge volumes but really about the certainty you get from looking at the whole playing field,” Dr Letchford said.
“My No 1 tip for any business is if you want to start rolling out more advanced analytics or machine learning, get all your data together — it makes the world of difference.”
However, a successful big data investment ultimately boils down to a few points: the core objective and just how hands-on an organisation wants to be when it comes to building the databases.
Social web analytics technology company Digivizer’s co-founder Emma Lo Russo said building things from scratch wasn’t always the best path forward.
“No one in Australia should really be building their own databases,” she said. “We’re a software company, we’re dealing with big data and I would still say Amazon Web Services, Google and Microsoft can do that a lot better, so an organisation can just focus on the bit that makes it different.
—
Originally published on The Australian.