Talking Data: Retail success online with e-commerce guru Steve Tadelis
If the Australian online retail market has any inefficiencies, companies like Amazon will expose them, according to Steve Tadelis, an adviser to the e-commerce giant, University of California Berkeley economics professor and Distinguished Fellow at Melbourne Business School’s Centre for Business Analytics.
Speaking at the Centre’s recent Talking Data breakfast, Tadelis said Amazon’s arrival in Australia, one of the world’s most popular online markets, will be a test for everyone involved in local e-commerce – from retailers to suppliers to the logistics companies in between.
“If the Australian market is very efficient, with high quality and low prices, then Amazon won’t make much of a dent,” Tadelis said. “But if it’s inefficient because of poor competition or supply chains, then it could have a big impact because it’s going to make people very happy.”
The happy people Tadelis has in mind are the many customers all over the world who have helped Amazon grow into a company worth around $US557 billion on the share market.
“Many sellers sell on Amazon because a lot of buyers like to buy on Amazon. It’s a place where they feel safe and offers them a level of trust,” he said. “The primary reason is because one of Amazon’s most important leadership principles is customer obsession – it’s all about delighting the customer.”
Building online trust was the topic of Tadelis’s breakfast talk, where he outlined how many of the most successful online marketplaces, including eBay for whom he once worked, achieve the right balance between full and no control over customer feedback.
“The biggest mistake is to rely solely on feedback that is user generated,” he said. “The platform or marketplace has to intervene with sophisticated data science to find signals in the data about sellers and the quality of their performance, and combine that with more market-based approaches that allow sellers to promote themselves in ways that favour the high-quality performers.”
Tadelis said finding the right balance involves a portfolio of activities rather than just letting people leave feedback in the hope that everything will then take care of itself.
“It’s never one thing or the other. You’ve got to find that delicate balance, and for different marketplaces, it might be a different delicate balance.”
Finding the right balance also requires creating well-balanced teams with the right skills to collect and analyse data and understand human behaviour and responses.
“If you put all your eggs in the machine-learning basket, then you’ll do prediction like nobody else, but that’s all you’ll do. You also need economists who can help you answer questions like how is behaviour going to respond to changes in the online environment.”
Tadelis says Amazon is now one of the largest employers of economists in the world, with Facebook, Uber, and other online companies following suit.
“Economists should be classified as another form of data scientist, with expertise in a slightly different scientific discipline. These companies recognise that when you’re talking about growth, product proliferation and innovation, vis a vis, customers, it’s critical to have people who are dedicated to answering those the kinds of questions.”
Our next Talking Data event on 5 December is sold out. It features global technology leader Dr Hugh Williams, another Centre for Business Analytics Distinguished Fellow, who served in senior roles at eBay, Microsoft, Tinder, Pivotal and Google Maps. You can listen to his recent podcast on Lessons from Silicon Valley: Success is in the data.