RETAILING is an industry where operating margins are slim and operational excellence is key to sustaining a profit. Nobody understands this better than NTUC FairPrice and in its constant push to achieving operational excellence, the tool of choice is information technology (IT).
Bernard Chew, NTUC FairPrice's CIO (chief information officer), notes that the success of the company's supply chain, inventory management, procurement as well as retail execution at its stores depends on IT.
"Our IT infrastructure gives our business leaders the visibility into the health and well-being of the business - whether the various parts of the ecosystem are running as they should be; whether planners are getting the information they need to derive the insights needed to the sort of long-range strategic thinking critical to our continued success."
NTUC FairPrice Co-operative Ltd was founded by the labour movement in 1973, with a social mission to moderate the cost of living in Singapore. From one supermarket, it has grown to become Singapore's largest retailer serving about 430,000 shoppers daily, with a network of over 120 outlets, comprising FairPrice supermarkets, FairPrice Finest and FairPrice Xtra.
Its convenience stores, FairPrice Xpress and Cheers, have a network spanning over 160 convenience stores islandwide and serves over 100,000 customers daily. NTUC FairPrice also owns a Fresh Food Distribution Centre and a centralised warehousing and distribution company.
Mr Chew notes that from an operational point of view, the key issue facing most local retail industry operators is the tight labour situation.
"Over the years, the industry has gotten used to cheap and abundant labour but now that has changed, and we are being forced to change and change quickly," said Mr Chew.
"It's not an easy thing to do, as retail and retail logistics is a manpower intensive industry. You need people to move products around, stack the shelves, prepare the food and produce and so on."
He adds that his organisation is using IT more creatively to automate traditionally labour intensive tasks in order to get the same work done with fewer people. He gives the example of the use of self-checkout machines, where one employee can now serve between four to six customers. Another example is the use of electronic price labels, where there is no longer a need to physically replace paper labels whenever there are price changes.
"Ultimately, it's good for our employees and the industry, as staff will get to earn more since they are more productive, and for the industry, there is an overall saving on manpower cost." He, however adds a note of caution: "Getting to this happy situation remains a challenge. Change is always a challenge."
Talking about the impact of major technology trends such as social media, mobility and cloud computing, Mr Chew recalls that FairPrice started out on its mobility journey by providing cellphones to key managers to help them keep in touch with the company - the so-called "Blackberry generation".
He recalls that during that time, Internet access was regulated for employees. "We wouldn't want staff to access social networks during working hours," he notes with a smile.
Now however, Internet access is more or less freely given, and why not? Today, staff already access the Internet through their own mobile devices anyway, Mr Chew adds. So why put obstacles in their path and make them less productive? "Google is a great information resource. Block access to that and you severely hamper your staff's access to useful, critical information and resource for continual learning."
He adds that security protocols that the company now has in place are more discrete and indirect. "We don't block, but we monitor. If employees wishes to access their corporate e-mail through their mobile devices, we require that they install our mobile device management software so corporate data can be remotely wiped off if the device is lost or if they leave the organisation."
"Yes, employees may be able to forward confidential documents or e-mails from their device. But if they really want to be dishonest, there are other ways to steal corporate information, like taking screenshots, walking away with hardcopy documents and so on. It's better that they access corporate information resources through approved channels where we can monitor or trace where necessary."
Mr Chew notes that social media can be hugely beneficial for enterprises. "The key in all this is customer engagement, whether in social media, or at our various touchpoints (like our online site FairPrice Online, as well as physical stores).
"We have had our own Facebook page for many years now and are active in responding to comments posted by our customers and engaging them whether about trending topics or feedback that we receive through this channel.
"This projects a more personal identity to an otherwise faceless corporation. At the same time, we also trawl the social media channels using automated tools for sentiment analysis to better understand our customers, their needs, wants and concerns. Overall, it helps us to better serve our target markets."
Mr Chew also notes that software analytics tools, combined with Big Data, are changing the way companies do business by giving unprecedented real time information about customers and also full visibility of the company's processes and supply chain.
There is now a lot more interest and understanding about what Big Data is capable of. "But insofar as making the most use of all the data that is being generated, we probably lag the leading markets like the US by three to five years," Mr Chew observes.
He adds: "To really appreciate the insights that Big Data can deliver, managers - not just corporate heads, but also line managers up and down the organisational chain - need to better understand statistics, which is the foundation of business analytics.
"And here, I think there is a big gap. Fortunately, there are a wealth of online learning resources like Coursera, Edx and Udacity, where suitably motivated individuals can go to brush up their skills and transform their careers."
From a FairPrice perspective, Mr Chew adds, analytics can help the company to improve inventory accuracy, putting the right products on the shelves, ordering the right quantities of each product, respond faster to changing tastes and habits of customers, their unspoken needs, better understand the different types of customers it has and address the needs of each segment - "the list goes on".
"We just started to build our data analytics capabilities a couple of years ago, so we are nowhere near where I think we need to be in terms of maturity in our data analytics capabilities. There are two pain points. One is educating managers in the organisation to the level where they can understand and appreciate Big Data and what it can do for them; and the other is getting data scientists who can create the data products that give business users the insights to make better business decisions or transform the business."
Getting the right talent is vital. You first need to get the right people in place who can ask the right questions that the data can answer, he notes, adding: "Then you need the data scientist who can mine the data and provide the data products that answer these questions. A good data scientist can discover insights that the business analyst wasn't even aware of - for example the correlations that were previously unknown."
As far as software is concerned, there are many tools available to mine even the largest data sets and present them through various platforms - whether traditional PCs on powerpoints, or onto mobile devices, he adds. Among the new and promising technologies is the Watson cognitive computing solution developed by IBM that would really come into its own if users can interact with it through natural language.
The Watson shot to fame by winning US$1 million in a Jeopardy match against two former human champions. It has two main areas of specialisation: one is Watson Medical and the other is Watson Engagement Advisor.
In the medical field, IBM experts have worked with medical researchers in the US in using Watson in the area of cancer research and medication. This includes collaborations with Memorial Sloan-Kettering Cancer Centre, WellPoint; the University of Texas MD Anderson Cancer Centre; and Cleveland Clinic Lerner College of Medicine of Case Western Reserve University.
In May last year, IBM unveiled the Watson Engagement Advisor, that helps businesses deepen engagements with customers. IBM earlier this year announced that it will invest more than US$1 billion in the Watson Group, focusing on development and research and bringing cloud-delivered cognitive applications and services to the market. This will include US$100 million available for venture investments to support IBM's recently launched ecosystem of start-ups and businesses that are building a new class of cognitive apps powered by Watson, in the IBM Watson Developers Cloud.
Mr Chew notes that in the case of Watson and other similar solutions, what is most important is the man-machine interface. "Current generation of tools - SPSS, SAS, R, Hadoop or whatever - still require skilled data scientists, statisticians and a technical ecosystem to produce the insights that provide value to decision makers. Even Watson, in its current form, needs a skilled person to feed it with the right sort of data, suitably wrangled to remove outliers and meaningless noise, and people to query it and get it to spit out the correct data. Not quite Star Trek yet. But I think we are getting there."
All this is being made possible due to the dramatic increase in processing power that a new generation of computer chips provides. However, as Mr Chew notes, the semiconductor industry is fast approaching physical limitation with regard to how much smaller and yet more powerful these chips can be made.
"We seem to be hitting some limits on Moore's Law as we get into 14nm silicon devices. This should spur more creative thinking on how we can continue to improve the technology.
"In a way, that's good. The past few decades have made IT professionals sloppy in terms of how we write programmes, as the hardware has tended to cover up for poorly written code. However, if hardware is no longer going to give us the order of magnitude performance increase every few years, then people will have to start using their grey matter again and learn how to write good, efficient programmes, which is not a bad thing at all."
Mr Chew adds that the thing to look out for over the next few years is machine learning systems, that is, systems that are able to understand data at the semantic level - in other words, the meaning of the data - and not just as bits and bytes.
"Another interesting area is how users can query the data using natural language. IBM and Google are doing a lot of interesting work in this area," he observes.
This series is brought to you by IBM