Effective use of big data in Hong Kong retail market

The other day, I was in a mall on Hong Kong side heading for a lunch meeting. Seemingly at random, I received a message from a store in the mall where I had shopped before, promoting their latest collection and with a buy three get one free e-voucher attached.

I am not a frequent shopper. I only visit this shop before every Christmas and yet I was triggered to drop by that shop immediately.

It got me thinking about the mechanism behind this intelligent marketing tactic and how the marketing team responsible had cleverly picked the right data, analysed, and then acted on about my purchasing patterns.

These days, the opportunities for big data are increasingly ubiquitous. But what are the most successful strategies for Hong Kong companies?

In one well-documented Hong Kong example, Ngong Ping 360 took Google's ad-targeting tool and integrated it with the company's own personalised marketing strategy. Once the system detected users who searched for terms matching the pre-set pool of keywords, the system displayed ads on the websites they visited.

This narrowed down the target market for Ngong Ping 360 users and identified potential visitors to Hong Kong at the same time.

Successful companies apply statistics

Every big data project begins with the need to identify exactly what we want to find out from the available data. And there is a lot available – big data is characterised by the 'three Vs' – Volume, Variety, and Velocity. Big data can be large in quantity, wide in variety, or very fast moving – or even all three – yielding many terabytes of information.

Statistics help us focus on two questions amid this deluge: which variables drive a higher profit and how can we trigger change in those variables? No matter how much data there is to analyse, researchers still need to apply basic scientific principles: creating a hypothesis, designing a test, and use the data to determine whether this hypothesis is true.

With big data, this 'truth' involves using cluster analysis to segment customers into groups with similar – hitherto undiscovered – attributes and then leveraging predictive modelling to predict behaviour and take targeted action based on these groups' known preferences and previous business transactions.

The deeper the insight, the greater the success

Big data is about much more than numbers and statistics though. Social analytics involves measuring the enormous amount of non-transactional – i.e. social – data that floods through servers every minute of the day.

Looking at social analytics means measuring awareness, engagement, and reach. “Likes” on Facebook, the number of times a video is viewed, the number of followers a particular page has…all these are big data gold.

This information can be used to perform targeted actions like customised marketing messages or cross-selling to particular groups. It can also be used to gauge the success of social media campaigns and other social marketing activities, allowing managers to tweak and adjust variables if a campaign is underperforming.

One example is Sony Mobile's efforts to leverage 'social influencers' – people who influence the choices and purchases of their social network. Their use of social analytics involves identifying these influencers through big data and then tailoring marketing campaigns to harness their natural influence and 'get their friends to buy along with them'.

Big data provides deep insight and understanding into customer behaviour, and creates a raft of opportunities for creative marketing. This type of knowledge carries with it a great deal of power – the power to see what a customer wants, and the power to win their hearts and minds.