Published on the 15/08/2016 | Written by Donovan Jackson
Local adoption lags, ‘big data’ a canard…
Soltius data scientist Andrew Peterson told iStart that he firmly believes New Zealand is a good decade behind the rest of the developed world in adoption of predictive analytics solutions. That’s a problem, he said, because of the power it has to transform existing and enable new business processes; just last week, Juniper Research provided an example of how the lending business is set for a nearly 1000 percent performance boost thanks to this technology.
“Predictive analytics is supposed to be really hot right now, but there is a generic national psychological roadblock that is preventing many of our organisations from even considering these solutions,” Peterson said.
The problem, in his view, is that there is a mindset that predictive analytics is ‘hard and expensive’ and so isn’t worth considering. Moreover, he said ‘big data’ is muddying the waters.
“There are a couple of details that need to be clear when it comes to analytics. Whilst it is all well and good collecting data, it is the analytics that brings it to life and creates business value. And the whole ‘big data’ phenomenon is, for just about all local businesses, completely irrelevant and unnecessary,” Peterson charged.
In fact, he believes the ‘big data’ hype is serving as a distraction and an obstacle which is causing something of an ‘analysis paralysis’, in that a good deal of companies are wrestling with just what to make of it. “There is too much emphasis on data and this notion that you need a lot of it. That’s just not the case; there are few, if any, companies in New Zealand right now doing true ‘big data’ in the way that massive internet companies like Netflix and eBay do,” Peterson added.
Having popped a few bubbles with those forthright statements, he said that real value can be found in ‘normal’ data. “What the vast majority of our companies need to get to grips with is that data by itself provides no answers, but the data you already have could make a real difference if it’s looked at in the right ways’.”
Where the magic is to be found is in the tools and the commitment to dig that value out. “Don’t get hung up about collecting data. Start with what you have, that will provide the insights on what you’re missing and deliver ideas on what else is possible. There are so many lost opportunities as a lot of organisations are getting caught in the marketing hype and can’t see the wood for the trees. There’s a lot of clouding that is preventing the typical business from doing smarter business.”
His advice is to approach analytics as an element of the organisation’s strategic plan. “Identify specific business objectives from the strategic plan, and use those objectives to drive investigation into advanced analytics. That way you will minimise the risk of failing. You can’t go bottom up ‘looking for insights in masses of data’ because it just doesn’t work for the vast majority of business. You need to go top down – starting with the strategic plan and working through the business objectives. And if you need assistance with that, then it’s time to get a data scientist in.”
Peterson makes another solid point. “Don’t get excited about the technology, get excited by how it can solve the problem. You don’t need Hadoop clusters and big data dreams. A standard SQL database will be perfectly adequate for most businesses. Add the right analytics tools, one or two good analysts and you’ll be set to go.”