EITN Roundtable Report Part 2: The Local State of Analytics

In all the time that analytics has been around, the business IT world has broadly categorised it into being descriptive, predictive and prescriptive.

It has been fairly easy transitioning from the descriptive type to predictive type, partly because besides making the world smaller, technology has been very adept at collecting data and storing it.

It used to be that We Don’t What We Don’t Know, but its scope is (slowly) decreasing, at least.

 

 

However, for prescriptive analytics to happen, data has to be truly insightful. For that to happen, data needs to be truly optimised.  This comes with challenges. The rule of thumb is that the more data there is for analytics systems to ‘massage’, the more informed the ensuing decision is likely to be. There is opportunity now, with Big Data or unstructured data. But this comes with another set of challenges, as well.

 

The two-fold challenge

Sarabjeet Singh, director of professional services at SAS Malaysia shared that organisations realise the importance of ‘clean’ data – ensuring that there are no repetitions, redundancies or inaccuracies. He added there is need for proper information management and processes from acquisition of data all the way to how it can be kept relevant over time.

 

“A common question is who owns the data and who owns the systems – in a lot of organisations, these were never documented and there had to be a lot of assumptions; in cases they were documented, they pointed to partners who didn’t exist anymore! That’s the reality of environments today. It’s a real challenge.”

 

So, businesses are still (unsuccessfully) ‘cleaning’ up data and now they also have to deal with a bigger than ever flood which; to make matters even worse; are in unstructured formats that traditional databases aren’t yet equipped to handle. Unstructured or Big Data accounts for 80-percent of all the data in the world, currently.

 

Charles Manuel, IBM Software Group’s business analytics executive for ASEAN said, “There needs to be an information agenda for things like governance, access, policy… you can have great sophistication at the top, but if all the plumbing is not working, you are going to have a stink.”

 

FICO Malaysia’s country manager Dinesh Pereira had said, “To do better analytics is to have all the data available.” But how do we not be overwhelmed by existing data that is ‘polluted’ and the exponential surge of new data? How do we optimise them?

 

Optimising data

There is a real lack of skills and experience in the analytics discipline. Andy Khoo, Netapp Malaysia country manager said, “The issue for cutting edge technology is a double-sided coin. If we do not hire fresh grads, how do they gain the experience which the industry is looking for?”

 

Dinesh Pereira FICO Malaysia’s country manager also made a case against the lack of analytics in customer service; the spray-and-pray mentality very much exists in situations where KPIs take precedence over almost everything else like customer satisfaction; a telemarketing agent calling as many people as they can, in the hopes that the number of people who will bite, will be more.

 

Pereira said, “The issue is with the way analytics is being used. For example, banks run into problems of annoying customers because advanced analytics isn’t in play yet.”  This is a crying shame because advanced analytics can mean all the difference between losing customers and keeping them loyal.

 

Analytics in the real world

The lack of strategic alignment within the organisation is also limiting real benefits. Khoo gave the very apt example of a car company that could tell there was very little demand for its product, but problems arose anyway because another division had targets to meet, and kept churning out more cars.

 

There were also examples about the use of analytics in the manufacturing sector. Manuel said, “The last couple of decades has seen a lot of process automation… but this isn’t necessarily optimisation.”

 

Sarab also added, “Manufacturing is also going into operations research, looking at things holistically, and also preventative maintenance. It brings so many disciplines together – different engineers who understand pressure, temperature, viscosity and all the other sensors that are out there today.”

 

This raises a huge think about how organisations could approach and implement analytics. Khoo thinks analytics should be thought of in terms of teams; a group of people with sound mathematical knowledge that is led by a business leader to give data a real-world flavour. Pereira had also shared that his organisation practices group knowledge transfer and Manuel alluded to the analytics quotient culture that ideally pervades across all levels of an organisation.

 

Other speakers of the Enterprise IT News roundtable also felt that centres of excellence (COE) and people of influence are key to driving analytics. Even empowering employees at the grassroots level is important because it is usually the people who are hands-on who are able to ask the pertinent questions that can answer the issues in a business.

 

Refer to part 1




Leave a Reply

Please Login to comment
  Subscribe  
Notify of