You know that Big Data must be near the zenith of its hype cycle when people refer to it as ‘the new gold’.
Though a touch hyperbolic, the goldmine analogy is a good one for conceptualising how every business is gradually waking up to an untapped resource on its doorstep. A bit like realising that the pile of rubbish you throw out every day is worth a million pounds per tonne.
That’s because it’s hard for people to think about data as anything other than a by-product. If we stopped and thought about data logically, it’d be immediately obvious why it is so valuable; why it needs to be protected and exploited.
It all starts with how you collect data and structure your processes
What use is operational intelligence that doesn’t tell you exactly what you need to know? Many organisations think that they have big data analysis when in fact they haven’t.
The approach will often be to give overstretched IT departments the challenge of creating a human-readable dashboard out of petabytes of information. But unless the foundations of data science are in place, you haven’t taken the time to understand the decision making requirements from the outset. So you can’t work back from there to determine which data should be captured, and how. Thus your solution becomes a sticking plaster of pie charts and tables that attempt to make sense of data based on its existing parameters and systems and, in so doing, pulls up short of your expectations.
The acid-test for these analytics comes when you try to drill a layer deeper into the findings and get nowhere fast. Good data strategy, where the integration of data sources and processes has been aligned to the needs of the business, will not experience such limitations.
A good strategy for big data analytics relies upon:
Engagement across the organisation. Business managers need to understand what value they want from big data analytics so that they can empower their IT team/data analysts. There will be constraints and obstacles and the business may need to be patient in having its expectations met.
Creativity and resourcefulness in data collection and retention.
There will invariably be critical data that isn’t being collected, or inefficiencies in using that data. Even core, business-critical systems like SAP can’t really drive an analytics dashboard, and present significant challenges getting data out of. All forms of internal and external third-party sources of data should be on the table.
A common platform for real-time data.
Information comes in many different forms from highly regimented database tables to the unstructured makeup of media files. This necessitates a common data model so that data can be mapped consistently across the business. This also maximises the visualisation capabilities of any subsequent analytics.
Only 4% of organisations are geared up to maximum big data returns
As if to prove the points above, according to a PriceWaterhouse Coopers study of 1,800 mid-sized and large enterprises in Europe and North America, 75% claimed to be “making the most of their information assets.” On closer inspection, the report found that only 4% were set up for success, with 43% gaining “little tangible benefit” and a further 23% deriving “no benefit whatsoever.” To be successful, according to the report’s authors, “the first step is to identify data sources, then understand the importance of analytics to every department and, finally, create a plan to stay competitive.”
A more recent report co-written by the Economist Intelligence Unit shows a similarly bleak picture of organisations reaching for the stars on big data analytics, only to end up floundering on the horizon. It concluded on three principal reasons for the shortfall between expectation and delivery, none of which centred on necessarily ‘spending more money on the problem’:
- Organisational issues caused by poor collaboration between business leaders and the people with technical responsibility for data analysis.
- Lack of a procedural implementation for analytics and of an analytical culture. The report identified that leading organisations had made investments in data science but that some of the biggest challenges to getting a return on that investment were “at the front and back ends of the analytics value chain.”
- Lack of integration between the various sources, systems and processes. According to the report’s author, “while over 90% have made the decision to move down (the) path to augmented, cloud based infrastructure, only 8% have connected the pipes with (their) analytic capabilities. This… gap really highlights how we’ve been building the plumbing but haven’t yet turned on the water to pull it through to the parts of the business where it can have the most impact.”
Ignoring the foundations of data science, and instead machining vast swathes of data to produce colourful, visually appealing ‘big data analysis’, is like drawing speech bubbles onto a photograph and calling it a graphic novel.
To my mind, data is central to everything. It’s the life of the business. In a roundabout way, that makes it a by-product too. John Lennon summed it up beautifully when he wrote: “Life is what happens when you’re busy making other plans.” You could say the same about data.
But data needs a plan, just like life. Planning it means you can choose to be impulsive, and react to anything with more confidence.