How Singapore’s NTUC Enterprise is embracing digital transformation

Guiding principles

What are some of the guiding principles you have adopted so far in transforming NTUC Enterprise from a traditional social enterprise to one that is increasingly digitalised?

Wong: I joined NTUC Enterprise to help with the digital transformation journey that had started by the time I arrived. One of the first things I had to do was to essentially answer a very basic question, which is, what does it mean to digitally transform the organisation?

To us, it is to become much more like a technology organisation that scales its business through the use of technology to expand into new areas and create digitally-driven business models – beyond digitising our processes. These could be building e-commerce capabilities, delivering an omnichannel customer experience or leveraging the power of crowds.

Second, digital transformation is about developing our own capabilities. For too long, we have been depending on external vendors and while that is still required, there are certain critical areas where it is best that we own the capabilities. That’s when we start to resemble a technology company where we have our own software development teams, designers and data scientists who can help us to innovate faster.

From an infrastructure perspective, moving away from legacy or monolithic systems to more modern architectures, whether it is cloud or microservices, is part and parcel of what a modern technology organisation should do.

Third is cultural change. Not only do we need the strategy and capabilities, we also need to have a different culture. It is highly admirable that NTUC Enterprise is both commercially driven and has a social mission at time same time. But how do we instil an agile culture within this operating framework? We want to do this in a way so that technology, data and new business models can be harnessed at a faster rate.

This requires a mindset change, from treating technology as a back office function to bringing data, technology and new business models to the management table. This is not easy because it requires changes to job scopes and KPIs (key performance indicators) to enable closer collaboration across different teams.

Keeping everyone on the same page

What are some of these new business models that you talked about? And how do you keep every organisation in NTUC Enterprise on the same page when it comes to digital transformation?

Wong: I have more exposure to NTUC FairPrice so I can talk about that. One of the key areas facing significant disruption is online groceries, where FairPrice On is still not the number one player in the market.

There is growing realisation that online groceries, in its current form, is not a profitable business. Customers are used to getting deliveries for free, but the costs of picking and delivering groceries erodes margins significantly in an industry where margins are already quite thin.

The introduction of new business models is not to continue doing things the same way. How can we fix the business model in an industry where the best player is still unprofitable, not only in Singapore but also in the US and UK? It is not a sustainable proposition to have an unprofitable business.

At FairPrice, we’re introducing a number of things to lower the cost of delivering groceries, which are typically picked and delivered from a central warehouse. For a start, we’re tapping one of the key value propositions that we have as NTUC FairPrice, that is, our proximity to customers through our network of physical stores across the nation. Today, we have three stores from which groceries can be picked and delivered to customers within a given area. This not only reduces delivery costs, but also improves the customer experience.

If you live within a three to five kilometre radius from those stores, you can get your deliveries the same day and sometimes as soon as two hours. And you can be assured that the fresh produce you buy is fresh because the products are picked from the store.

We’ve seen some good results in terms of economics and customer experience. We still have a central warehouse, because that’s the world where we’ve come from, but we’re also increasingly looking at the store fulfilment model.

Scalability

Well, with a central warehouse you can automate warehouse functions and use robots to do the picking, which is what many e-commerce players already do. In FairPrice stores, I don’t think you have robots to pick products, right? So how scalable is the store fulfilment model because one of the things you talked about is scalability? Do the stores have enough people to go around picking items while attending to walk-in customers at the same time?

Wong: Yes, you're right. With a central warehouse, you can have machines picking items and that helps with costs. However, in the store, you have to find a balance between providing access to fresh produce and offering a wide variety of groceries.

We’ve been doing groceries for three to four decades, and we already have the refrigeration systems and processes in place to identify what needs to be chilled or frozen in the stores. But to do that in a central warehouse takes a lot more work and investment. You have to track product expiration dates and maintain quality levels. We can do that, but it’s much more costly.

In our central warehouse, we are able to keep a wide assortment of groceries of about 20,000 stock keeping units (SKUs), but in some hyper marts that we’re delivering from right now, we have been able to increase the assortment significantly to 40,000 SKUs.

Now, you may say let’s also increase the assortment in the central warehouse, but today the space is constrained. The benefit of delivering from a store is that a store’s inventory can be optimised to serve the needs of people living in a certain area over time. So not only do we have a bigger assortment of products in some stores, the assortments are also tried and tested.

As for automation, we’re introducing picking technologies at the store to improve efficiency. Though they are not as efficient as a robot yet, we’re thinking of how to get there. For now, we have an app that our picking associates use in-store to guide them on the fastest routes they can take to pick groceries.

The store fulfilment model also gives us more flexibility in meeting seasonal demands. While efficiency is important, a tad more important is flexibility. For example, if we anticipate a lot more orders from a store tomorrow or in a month’s time, then we can hire additional people.

But if you’re making capital investments in automating a central warehouse, then you can’t magically pick 3,000 more orders per day. That requires additional investments which will not be utilised after the peak periods, whereas in the stores we can have a more elastic workforce to handle those peaks. That said, I think we’re going to have both models for a variety of reasons and we will be making improvements in both models.

Consolidating data

What else are you doing at the back end to support these digitalisation efforts?

Wong: We’re doing a ton of stuff on the back end. For example, we are consolidating data and leveraging information across FairPrice stores to improve personalisation for customers. When we send e-mails to customers, we can include information about relevant promotions and products that customers tend to buy, rather than some random promotion that may not appeal to them. While these things are still what I would consider very much part and parcel, we’re getting better. There's a lot more exciting stuff coming up.

What about optimising your supply chain? People talk a lot about the use of the internet of things (IoT) to track goods in real time and some food companies are also experimenting with the use of blockchain to trace the provenance of their products.

Wong: I would say we're in the early stages of exploring IoT when it comes to groceries. The best of the best are trying to use blockchain technology to track the provenance and handling of goods by different parties, but the effort requires a lot of collaboration across all industry players – from the people who grow bananas, for example, to the shipping companies. I think we’re still in the early days and I’m watching this from a technology perspective. But I can also see that it is very difficult to get that coordination going.

You’ve talked a fair bit about FairPrice – what about digital transformation efforts that are going on in other businesses like the NTUC First Campus chain of pre-schools?

Wong: Indeed, we’ve also done a lot of work for First Campus, including things you don’t see as a parent. For instance, we’ve helped to automate and digitise a lot of processes like registration, which used to involve a good amount of physical paperwork.

When it comes down to the interactions, we are already enabling apps and web interfaces that parents can connect to find out what’s going on with their children. We also provide feedback to parents on how their children are maturing and what additional things they can do.

Supporting digital transformation

What were some of the changes that were made to support digital transformation from both organisational and technology infrastructure perspectives?

Wong: For change to be effective and lasting, it needs to be inward. Therefore, all the innovation work that we’ve done is driven from within existing businesses, as opposed to other organisations that may create a digital lab or some other outfit that is separate from the business.

The latter approach provides more leeway for organisations to experiment without being held back by the constraints of a larger organisation, but when the rubber hits the road, which is in our business, you don’t get to scale.

So for us, it’s better to start the journey inward, which is why I am also the operating head and managing director of technology at FairPrice. I am driving these changes inwardly within FairPrice, as opposed to setting up a different organisation.

From an infrastructure perspective, moving away from legacy or monolithic systems to more modern architectures, whether it is cloud or microservices, is part and parcel of what a modern technology organisation should do. We are putting more workloads on the cloud, and for things that we are not differentiating ourselves by, we will leverage best-in-class services so we don’t have to be the ones operating our own e-mail server.

We also have development operations, or DevOps, to automatically push code into production, and to track the performance of new software releases. On the design side, we have a significant design team to look at problems with customer and user experience, such as how and where users are using our services. We’re also investing a lot in data science and data engineering. Big data was a buzzword a number of years ago, but it’s still very much relevant.

We're investing a lot in core product development, whether that is in front-end technologies, the web, which is still very much relevant, and of course mobile engineering. Our mobile engineers are backend engineers who specialise in creating application programming interfaces (APIs) and to make sure that our APIs can scale up and down in terms of usage. Our security practice is quite good too, because we know that if hackers want to make noise and get recognised, the best way is to target a name brand like NTUC.

Assessing the outcomes

How do you assess the outcomes of the digital transformation initiatives that you have described? What are some of the ballparks that you are looking at, say cost savings or customer satisfaction scores?

Wong: Yes, we look at cost savings, but we also look at productivity enhancements. So if it took 10 hours to do a particular task, and now it takes one hour, that’s an improvement. We also have KPIs on service availability, incident response times and the amount of revenue being driven or enabled directly by technology. These could be online revenues or revenues that come through our digital interfaces.

What would you say would be the biggest challenge you have encountered? You talked a little bit about some of the challenges, and I can sense that a large part lies in the cultural aspects of digital transformation.

Wong: One key challenge is certainly the mindset for transformation. The teams that I need to build, those I bring on board or have inherited, as well as the stakeholders that I interact with, all need to align better in terms of the speed of change and the patience for change. And by patience, I don't mean just time, but also potential risks, failures or delays. Instilling the mindset to stay the course, and not to be flustered, is what I need do better in.

That said, it’s really admirable that our senior management and board, and through them, a larger constituency of people, are for change. And for the transformation that we’re going through, it is nothing short of amazing.

We’ve gone through some big discussions about changing things and investing in things without seeing the same level of returns as other projects we would do. That’s nothing short of amazing, and I have to give it to their foresight and aspirations for impact. However, of course, there are people who have not been able to adjust to this mindset. Perhaps, people who don’t want to be suffering the change, whether they are from the existing organisation or the new people I bring in.

Sometimes, I bring in people who think they’re going to be in a green field setup. But sometimes if they’re not placed in these setups, they might be doing the work they want to do 30% of the time while the rest of their work is to convince others to do the right thing.

Adjusting to the emotions of a discussion

It is in that convincing where you need to have to have the mindset that not everybody’s on the same page, and you need to do a lot more work in explaining and adjusting to the emotions of a discussion.

Some people just cannot take that and say they weren’t paid to discuss and argue here. Well, it’s not about discussing and arguing but about educating – because people have been doing this for many years and have their own views. You can’t just come in one day to explain things and expect everyone to change.

Then, more generally speaking, with the digital economy, there are new mindsets that are emerging. For example, the big technology companies are willing to be unprofitable for a long time in order to gain market share. That is not our standard mindset, because traditionally we’ve been very profitable. And investing in things for which you don't see near-term gains is also a foreign concept.

Another one is the speed of execution. As you know, start-ups often release just 10% of what their first product could be to prove the value proposition of what they’re trying to do. They’re not worried about all the other things that would make it complete. That's also what I'm trying to do here.

Sometimes, we need to release functionality that is incomplete, and we may be missing things here and there. These things may be important, but there’s a more fundamental hypothesis that I need to validate early on.

Because if that hypothesis is not validated, even if I have a 100% complete feature set, the whole thing is not going to work. So having organisational acceptance that we’re going to do minimum viable products will also take some time.