11 May 2015
Data analytics: Distilling value from big, challenging data
There was a time when data was well behaved and gave up its secrets easily. Today it’s big and messy. We’re in the midst of a data explosion like no other and the question ‘what should we do with it all?’ has never been more important.
A tidal wave of new information from sources as diverse as web interactions, wireless sensors, mobile phones and now the Internet of Things could provide unheralded insight into anything from detailed customer behaviour to exactly when a piece of factory machinery is about to fail.
It holds the promise of cost efficiencies, growth in existing business, birth of new revenue streams and even monetisation of the data itself. However its sheer volume and variety defy normal analytical methods. That’s when data analytics takeover, a set of techniques and technologies that tease the meaningful from the massive.
So what do decision makers need to consider when wrestling knowledge and profit from the data they hold?
To begin with the computing power to crunch it is far beyond a desktop machine. The future for this analysis lies in the Cloud where spreading a problem across thousands of servers delivers answers fast enough to be useful without requiring esoteric hardware.
Firms that use engagement analytics experience a 7.6% boost to customer lifetime value.
(Source: Aberdeen Group, May 2014)
Then what? When David Stephenson was eBay’s head of global business analytics he ventured that the company’s colossal effort analysing 100 million hours of customer interaction every month had a particular ambition - to make the vast business appear like a small, local shop, one that knew its customers really well.
An ironic strategy but an important point as in the rush to become ‘insight driven businesses’ other organisations have been guilty of embracing data analytics without really knowing what to do with the results. Neil Mackin, analytics director at Capita suggests a few simple strategic cross checks before considering what data to gather, how much, how often (and the partner you might need):
Customer engagement analytics users enjoy 79% greater annual improvement in client satisfaction rates.
(Source: IBM Center for Applied Insights, October 2014)
- What benefit will the work create (increased efficiency or effectiveness, better cash flow, new business)?
- How might you go about the analytics (data sources, data science methods, skills, tools, etc)?
- What needs to happen for the benefit to be realised?
- Can the people within the business do what needs to be done?
The analytics and technologies themselves often need to respond very rapidly to be worthwhile – for example fast enough to intervene in online fraud detection. Success may require people across an organisation to work in new and unfamiliar collaborations.
Despite the potential some organisations remain nervous. Issues around data protection can be complex and the public has yet to be engaged in a debate on what should and what should not be done with the personal data collected about us. Nonetheless if data analytics has proved anything from the likes of Netflix or ASOS and more, it’s that it has the power to radically disrupt existing business models. This is unlikely to be a wave you can afford not to ride.