Don't Wait to Tackle Big Data (or Risk Falling Behind)

Still considering a big data strategy? Don't wait too long â¦ your competition is surging ahead based on predictive modelling power

Big data is no longer as scary as it once was, and as a result, over the last few years, organizations around the world have been scrambling to engage it, even without a clear strategy. One fallout is that companies are literally drowning in big data, explains Alessio Marinelli, with much of the internal data still stuck in functional silos.

Marinelli has lived and breathed data for over 15 years, first as KPMG's Data Analytics and Business Intelligence lead in China, and now as Director of Analytics at Advanced Merchant Payments, based in Hong Kong, where he devises models that leverage big data to more accurately evaluate the creditworthiness of small businesses of and minimize the company's risk exposure for its loan portfolio.

[eventpdf]

3 Tips to Devising A Big Data Strategy When mining big data, it's not the quantity but rather the context that counts, Marinelli explains. Most companies start out by amassing as much data as possible, and end up overwhelmed. Instead, companies should be asking themselves three things as they devise their big data strategy:

Do I have a clear vision of my business strategy? (Ask the right questions) Do I understand the information requirements to support that business strategy? (Leverage the data you do have) What additional data requirements do I have outside my internal data, which should drive my big data strategy? (How to complement internal data with enriched big data)

Taking a step-by-step approach to a big data strategy avoids the all too common scenario of companies marooned in data without a structured framework on which to build their analytics.

Asia: Still Waiting? While a number of commentators have described Asia as "sitting on the fence" with regards to big data, the signs are clear that some organizations, notably Chinese, have learned the lesson very well.

"If you think about the success of Ali Baba, or WeChat, or even banks like CCB or ICBC, you can see that they have had massive returns on their big data investments" says Marinelli.

The Economist Intelligence Unit, in a survey of more than 500 executives across Asia Pacific last year, highlighted the limited success of big data strategies so far [more than 50% of respondents claimed their progress was limited] whereby the biggest hurdles cited were internal, mainly relating to difficulties sharing information across organizational boundaries.

The report cites 4 defining characteristics regarding the uptake of big data in the Asia Pacific region:

Adoption is slower then expected Internal issues are inhibiting adoption Frontline employees cannot access the data they need There is a strong belief in the benefits of big data.

The Dangers Of Being Stranded Behind The Big Data Wall For those companies still standing on the sideline, uncertain as to how to proceed: There is a cost associated with waiting. And according to Marinelli, itâs significant.

"While we have plenty of examples of companies who have experienced extraordinary growth as a direct result of combining internal data with relevant big data to identify a more profitable customer segment, for example, or growth markets â we can also see how the absence of a big data strategy is causing other companies to fall behind," he says.

With competition now global, and the playing field levelled, the danger of giving up a big data advantage to the competition is simply too great to ignore, he says.

Big Data for the Back Office While the obvious examples of big data's impact tend to focus on better-servicing and retaining end customers or users, this overlooks â all too easily â the opportunities presented for internal processes like Finance.

"Companies expose themselves to sub optimal models if they are still relying on year-over-year statistics, as opposed to big data," explains Marinelli.

"For example, if you take the procure-to-pay process, the precise predictive models incorporating big data can more accurately forecast demand, with the knock-on effect of more accurately forecasting the quantity and quality of products required for procurement. Big data, therefore, naturally supports the trend towards increased collaboration between a business and its suppliers."

Companies not availing themselves of this kind of modelling run the risk of building up unnecessary stockpiles and tying up valuable cash, he warns.

Similarly, for organizations operating across multiple countries, big data enables improved modelling for transfer pricing, for example, or currency buying, by more accurately evaluating various external conditions as a result of including flexible variables (namely, variables driven by big data).

"Obviously, the more accurate your models, the more closely you can hedge yourself against a more predicted future," says Marinelli. Companies still relying on traditional models risk over-hedging or under-hedging their positions, and are in a less favourable situation compared to more proactive competitors.

The Time Is Now Big data is complex, but providers have stepped up over the past few years to offer data mining and data analytics services, including scraping social media and other external sources for big data, which go some way to removing the complexity for organizations still uncertain as to how to move forward.

The challenge is that time is moving swiftly, and every moment a competitor has to gain an inch of data-derived advantage will cost those who are lagging behind.

Asian companies cannot rely on the advantage of labour arbitrage forever. Big data is a significant tool in the "performance" toolbox, and has the capability of driving advantages through most every horizontal and vertical business process.

"Whether you're applying that capability to identifying risky customers, unreliable suppliers, or any other aspect of your finance function, you simply cannot leave the competitive advantage on the table for others," says Marinelli.

Organizations that have not yet incorporated big data will find that they are no longer competing on a level playing field. Competitors who have embraced big data already are able to leverage predictive variables more accurately within their future modelling scenarios, and thereby minimize their use of scarce resources and control their risk exposure.