The latest insights from your peers on the latest in Enterprise IT, straight to your inbox.
ASB, a New Zealand bank, takes a practical view of big data — an approach worth noting, given the nation’s embrace of new technology and its mobile, social-media-savvy consumers.
By Peter Gavin, Head of Data, ASB
Thinking about investing in big data or perhaps worried about whether you should? If this is you, then read on!
Big data has captured a great deal of attention in recent years and has spawned much confusion. While the physical management and processing of large volumes of data gets a lot of airplay, more-fundamental questions relating to the nature and general value of data and, crucially, its application to business deserve more attention.
And I have a somewhat special perspective on this.
At ASB, we have developed an approach that closely considers the needs of customers and the business, as well as the most effective use of IT resources.
In 1969, we became the first New Zealand bank to operate an online real-time banking system linking all our branches. Since then, we have continued to actively pursue new technologies where we see the potential for value. This process has been assisted by New Zealanders, who embrace new technology, are highly mobile, and highly oriented toward social media. Indeed, we hear that many of the global social platform leaders test new ideas in New Zealand.
A Three-Dimensional View
At ASB, we think about data across three dimensions and believe that showing how big data applies across these illustrates an important message about the topic of big data itself.
1. Managing data. From a big data perspective, there is a greater variety and volume of data than ever before. There is also more complexity in gathering, retaining, and governing that data while balancing a need for increased security with greater customer accessibility. As custodians of our customers’ data, we take a no-compromises approach to ensuring its security and appropriate use.
2. Extracting insights. New and cheaper analytical techniques and tools have allowed access to new parts of the data “estate,” and new communities of interest have spawned around data at different parts of its life cycle — for example, Marketing is interested in what is happening right now in oursocial channels.
3. Creating value. Our organization has invested heavily in a “Sales and Service” culture. We are excited about what this means in a future where our customer interactions are increasingly digital, “social-aware,” and driven by data. The greatest challenge in all this is how you take data and make conversations or digital experiences more relevant and valuable for customers. One of our executives describes this as figuring out how to balance insights and instincts, which I think perfectly captures the opportunity and the challenge.
Vendors often focus on the first and second of these dimensions, but few talk about the challenge of creating value from the data. Achieving this really depends on changing the culture of the organization to become more data driven, so that it can help customers leverage their own financial data. We have recognized this problem at ASB, and we have learned that any kind of big data or business intelligence, or BI, activity should start with a question. Asking and answering the right one unlocks real value for our customers and shareholders.
Asking the Right Question
An effective BI strategy always involves trying to determine the business question that needs an answer. When we have that it is much easier for the technical people or a vendor to deliver the right data. In fact, if you don’t have that right question, the work being done may be simply academic in nature, producing interesting information that is of little value. We try to define information that the business can use, information that will allow its possessor to make better decisions and “turn a dial” to increase customer satisfaction or market share. When that happens, information really does have value. The question asked was the one needed to focus analytical resources efficiently.
In some cases, gains can be realized by employing the art of the possible – that is, using our growing capability to analyze big data to see what sort of insights emerge, to identify potential areas of opportunity or discover patterns in the data that would not typically be noticed through human intuition. I have a small innovation budget that supports low-cost and low-effort initiatives to provide a quick view of topical data or seek out interesting patterns.
I’ll take the findings to the business, and if we’re told that something is not of interest, we move on to the next idea and the next view of the data. On the other hand, if they say, “This is interesting and could perhaps be even more interesting with certain additions or refinements,” then we can make quick gains and focus on how value could follow from answering a specific question based on a perceived opportunity.
The Future of Big Data
What is happening with the evolution of big data is not new. There are many parallels with the evolution of BI strategy in the 1990s. What we are experiencing now I believe is really BI 2.0, with new tools, data, and potential revenue streams based on our increasingly digital lives.
BI, which analyzes structured information in relational database management systems, typically routes data through a data warehouse and yields insights the next day. It is easy to interpret and analyze because it is highly structured — but that represents only a small percentage of the data within an organization’s environment. In recent years, it has become economically feasible to analyze the next, say, 35 percent of a company’s data. For example, it is now possible to analyze documents, e-mails, and social feeds using inexpensive tools that can reveal new opportunities to help customers manage their finances and alert us quickly to their needs.
Still, even in BI 2.0, at least 50 percent of data remains untouched — and some of that isn’t even “data” yet. As an example, a casual conversation is currently classified as ephemeral data, but in the next few years I believe we will see more advanced voice recognition functionality on devices like smartphones, allowing conversations to be analyzed just as social media data is today.
If you were to correlate those segments of data to revenue availability, you would find that a lot of the easy-to-grab revenue is in the easy-to-analyze data (BI 1.0). If your company is comfortable that it has already tapped most of the revenue there, then it makes sense to start tapping into the next 15 to 20 percent or beyond. Perhaps only 5 percent of the value is in the last 50 percent of the data. And certainly the cost to extract value from that 50 percent is much higher, because the tools for doing that are not well developed. So, if you have extracted most of the value from BI 1.0, that’s great. But if you haven’t, I would recommend focusing your efforts on ensuring you are getting the maximum return from what you can do today.
At ASB, we first want to ensure we have the right foundation in place. We plan to exploit further growth in data while doing what’s needed to keep up with regulation and making sure we generate the right types of opportunity. And we never lose sight of the fact that the future of big data will involve considering big data as a complete ecosystem, one that involves not just the raw data but also people, processes, culture, and tools.
Originally published in CIO Straight Talk, No. 6 (February 2015)
There’s lots of talk about managing data and extracting insights from it, but not enough talk about the challenge of creating value from the data. Any big data or business intelligence activity should start with a business question that needs an answer. In fact, if you don’t have that question, the initiative may yield information that is interesting but irrelevant.
Although we have tools to analyze mounds of unstructured data, at least 50% of data remains untouched — think of casual conversations. The future of big data will involve using technology to mine those vast untapped sources , but companies can focus today on the easy-to-analyze data, which is currently generating the easy-to-grab revenue.