“But then we do move on to the technology, starting with the strategic IT planning to industrialise all of this applied analytics, placing it at the heart of all processes. We need to have the proper data management structures in place, which is indeed where the scale of Big Data can become a significant part of the challenge. So the business leadership has to be followed by skilled and professional IT resources. But the experience of those organisations that make a success of it is that their analytics investment continues to deliver ongoing improvements and opportunities and insights,” he says. “Even wringing, say, 0.5% of a gain out of any of your business activities will positively impact the bottom line.”
Integrating data
“Big Data is really just a buzzword media hype sort of term, I think,” says Dr Yasmeen Ahmad, Principal Data Scientist with Teradata and a lecturer in the subject in her alma mater the University of Dundee. “When it came into the ICT market the early stuff was all about social media like tweets and producing nice visuals and graphics from the massive Big Data resources. But in truth there was little real value.
“Our thinking has matured today and for us and our clients it is really just another data source to add to your repository of data with the possibility of enhancing your analytics and what you are achieving from that. It’s not about ‘the Big Data’ but about integrating all of your data and combining it in a disciplined and useful way.”
“In general the kinds of business organisations we deal with have been doing analytics for years across a spectrum. We see most of them as genuinely great at creating business intelligence, regular and accurate reporting of how the business has performed with as many criteria as they need or wish. That is valuable but is largely numbers reporting,” Dr Ahmad believes.
“Where we are moving forward is in helping them take a big leap into predictive analytics, using patterns and trends and insights from their historical data to successfully predict future outcomes. It can give them the information to change course with their customers, for example, not just predicting what they will or might do but actually intervening and positively changing that customer journey. With today’s technology that can be in real time interventions as customers are on a web site, for instance offering the assistance of a sales agent at a key moment.”
Data quality
One interesting point she makes is that when we are talking about Big Data we can take a slightly modified approach to the data quality. “When we were performing analytic on samples we had to be 100% sure of the quality of the data. In this new world where we are analysing all of the data together, for perhaps tens of millions of customers, we do not have to be as hung up on the data quality,” Dr Ahmad says. “We are more in a discovery analytics mode, so perhaps 80% quality might be good enough.
“When you come to operationalising what you have found you will go back to high quality requirements because you have a focus. There are new analytical techniques like graph or path or text analytics that can complement the traditional data science and statistical tools to leverage better information from a body of data.”
Clearly IBM has been in the data management business for a long time, and has a specialist team in its Irish operation working with some major customers on applied analytics in big data — and the traditionally enormous data stores of banks, insurance and state bodies plus some newer sectors. “I am particularly engaged for some time now with one of our larger banks and with a highly successful online Irish betting company,” said Jason Burns, Analytics Client Architect in IBM Ireland.
Early days
“It is still very early days for data analytics in general, and its application in different business sectors. It’s enormously interesting and challenging because, especially in the bank, it is so clearly opening up new possibilities across the board. Like many banks today, our clients are investing strongly in analytics and setting up specialist teams from different backgrounds. They constitute a mix of disciplines, notably statistics and data science, IT itself, actuaries and the business itself — which has many specialist strands of expertise and experience.
Ltd
