OCBC’s analytics strategy and what brands can learn from it

Analytics for OCBC originated as a small function on the consumer banking side, but over time was elevated to become a group function. All the direct marketing duties for the consumer bank came to the analytics team.

Last year, it ran 1,600 campaigns which led to 19 million leads, generating more than $85 million in shadow revenue for the bank. Almost 30% of credit cards and 35% of wealth management comes from leads generated by the team – so it’s a significant generator of revenue for the frontline.

What was the journey like? I sit down for a chat with Donald MacDonald (pictured), head of group customer analytics and decisioning for CRM at OCBC Bank.

MacDonald leads a team of nearly 40 called G-CAD (group customer analytics and decisioning) which he says acts like a centre of marketing analytics excellence in the OCBC Group. The team is broken into four divisions – relationship management, analytics (or the data geeks), campaign management and CRM – and supports the business’ banking side, Great Eastern, OCBC securities, and even human resources.

In my many conversations with marketers, a common challenge faced with analytics is that you have a lot of smart data guys sitting at the back, but they can’t translate the analytics into business action because they don’t speak the same language as the business.

One of the ways, MacDonald worked around that was creating a relationship management (RM) team. The team boasts people who sit between the data geeks and business managers and “translates” each others’ needs and also acts as a filter.

“If it’s just a pipe between the business and the analyst, there will be a deluge of requests. We only have limited capacity. So it has to be filtered. One of my biggest challenges is that I have too much work, not too little, so the RM is kind of a filter there,” he says.

The RM team is the one that centralises the learning as well.

“If you’re dealing with individual analysts, they may not have the bigger picture for each of the business units. So we’re probably the first team to introduce a model like this and I think it was one of our greatest successes actually,” he says.

The first steps: How it all started

OCBC had a basic analytics function even in the late 1990s. It had a warehouse, some campaign management tools and was doing data marketing, albeit pretty rudimentary. Like a lot of banks, it invested very heavily in CRM back around 2001.

Back then, MacDonald was a consultant, and consultants, typically, would go around to banks promising huge returns from CRM and so on.

“So OCBC implemented CRM, expecting these returns and they found the consultants hadn’t maybe told the truth. The returns they got weren’t what they expected. In 2001, they did an analysis to find out why the CRM investment did not really pay off,” he says.

MacDonald uses an analogy to explain. Back in the day, the bank built a great pipe to the customer; Siebel was a great platform and it had a well-trained salesforce, but “the oil that was going down the pipe was no good, i.e, the analytics at the back was not sophisticated enough”.

“So the quality of the leads was very poor to be honest.”

OCBC managed to build a business case to show the analytics could return some business value. In 2004, it built its core infrastructure – an enterprise warehouse, which allowed the bank to see every single customer, every single product, and every single transaction that was made.

That’s when MacDonald joined the company to run that project. OCBC integrated that data with all the channels with the focus to maximise investments.

The bank had invested $10 million and it had to pay the investment back. “With the returns, the incremental lift that we created, we managed to pay it back in just nine months. And it was actually a huge success which exceeded all of our expectations.”

Expansion: Getting the right talent on board

You can have great data, a great platform, but if you don’t have great people to take advantage of it, then you are bound to fail. What did OCBC do to hire the right people?

[For more top marketer perspectives on talent, look out for Marketing Magazine’s The Talent Forum on 12 November at Four Seasons Hotel. To register, please click here or contact Jason Chua at jasonc@marketing-interactive.com, +65 6423 0329.]

“We spent a lot of time then, strengthening the team and expanding. The first year when I arrived, I think we only had maybe six people at that point in time. So we had this great engine, but not enough people to make use of it,” he says.

OCBC then focused on maximising and strengthening the team and working with the business to make sure it made use of the improved data and analytics capabilities. That’s when it introduced the RM model as well.

From 2006-2008, it was all about maximising this in Singapore and given its success, it was asked to regionalise the effort. In 2008, OCBC effectively extended the capability to Malaysia and China, which are the bank’s other major operations. It was the same year, Macdonald set up its offshore team in China, in keeping with its regional plans.

Since then it has also taken on the function for the entire OCBC Group.

Shifting the focus to small data

Increasingly, the role of this team is becoming real-time. If a customer opens an account with OCBC, the bank shares the entire process with the customer during which it captures the information about the customer and passes it back to the data warehouse to see if it knows anything else about that customer.

Even if it doesn’t, it is still able to score the customer and present an offer, while they are sitting there with a salesperson.

In terms of the data, the analytics team works with the product and segment managers to let them understand who their customers are. If they can understand that, they can obviously develop solutions that meet the needs of those customers. So the most basic thing the G-CAD team does is customer profiles – from individual profiles to standard ones to customer DNAs where it is looking at different segments.

The team over the past nine years has had access to five billion transactions and a lot of those are credit card transactions. “We use big data to make the small things relevant. Small things meaning even the SMS you get on your phone.”

It has identified about a 100 different segments/usage of its credit card customer base. “We then use this to influence the SMSs they get from the card to make sure they’re relevant to the things you’ve actually bought. We also use this for strategic purposes such as the launch of Frank, OCBC’s youth banking arm.”

There is a lot of talk about big data, but OCBC gets a lot of power out of small data – and a great example is complaints.

“We probably only get about a 1000 complaints a month which I don’t feel is a lot. But from that very small data set, I can get a lot of good insight.”

Once the team has analysed the nature of the complaint, the profile of the customer who complained, the reasons behind it, the bank can understand what to do next. It also uses it to predict who is likely to complain in the future and what nature of complaints can be expected. Once it has found that insight, it needs to be shared.

“It’s not about having an analytics team sitting in the ivory tower, but one that is embedded in the business and actually changing things on a daily basis – and that only happens if you share the daily insights and information.”

The analytics team does that at different levels of the business. It comes up with net promoter scores for senior management looking at strategic segments, while also coming up with that score for the frontline staff, but sliced differently.

“If you’re the one picking up the phone, you may wonder how you are in comparison to your team – how is your satisfaction versus the other people on the team. We’d also look at tenure – we do this for the training team – so we look at net promoter scores for new staff versus experienced staff,” he says.

Increasingly, the learnings are also being shared externally.

“We use our analytics capabilities with our partners to try and deepen the relationship.”

A great example is Robinsons Group – one of the bank’s biggest credit card partners.

OCBC’s analytics team has helped Robinsons understand who its customers are and build a capability like theirs on an outsource basis. It also did similar work for Marks & Spencer. The partner gets increased access to insights and OCBC gets more access to data.

The way forward

One of the challenges for the analytics team is to manage enormous amounts of data and insights which goes through it. What it’s trying to do, therefore, is to deploy the power of insights into the hands of the business users so they don’t always have to come to the team.

The offering called QlikView allows the end product manager to do the analysis themselves.

“It is a very nicely built front-end – you go in there using pre-built reports and dice and slice the data up to the individual transactions.”

Unlike many other dashboards that are aggregated with a lot of predefined dimensions, this tool is a memory intelligence tool which is a lot more powerful, he explains.

“The user can literally go down to the individual transaction level. So if you’re the credit card product manager, you can see how the trend is going in food and beverage, or at outlets, say Thai Express, and so on.”

“It’s a lot more about empowering the user.”

And the next frontier is HR. Things such as recruitment and being able to predict which candidates are likely to stay for a long period of time.

“We’re basing that on people we’ve hired in the past. We spent millions of dollars on training, which training courses work, will you benefit, will your performance lift, who is the best trainer to take that cost, etc. These are the types of decisions that we can use data for to help us understand things.”