Let me provide you with some data. 100111010001.
There – that’s useful isn’t it? OK – it’s not; but that’s the problem with data. In itself, it has no value – it only starts to deliver value when it is analysed and contextualised.
There is the oft-used phrase of ‘drowning in data’, and, to paraphrase Coleridge, for many organisations it is now a case of ‘data, data everywhere, nor any stop to think’. All too often data is collected, but is then not used due to a perceived lack of capability in dealing with it in a satisfactory manner.
There is a need to deal with the ‘V’s of big data analysis – you may be dealing with high Volumes of a large Variety of data types, at high Velocity – with the Veracity and Value of the data being suspect.
The question then is – how does your organisation extract the value from that data that enables you to move up the strategic insight pyramid from a sea of data, through to a pool of information, on to a puddle of knowledge that gives the drop of knowledge that creates that strategic insight that adds to the organisation’s bottom line?
Effective analysis of that sea of data needs to be centred on knowing when and where an activity occurred. Having access to this data allows for trend analysis (using the time variable) and also for spatial analysis. By combining the two, predictive analysis can be brought to bear that can add massive value to an organisation.
Let’s consider a retail organisation. It has collected lots of data about its customers through the use of loyalty scheme cards. It therefore knows what its customers bought from it – and should also have when these items were bought, and where from. It should also have enough details about the customers – via their post codes, for example – to be able to position them on a map and assess their socio-economic status as well.
The retailer can therefore look at purchasing cycles of certain foodstuffs – just when do customers start buying strawberries – and do they buy cream at the same time? This basic analysis can avoid bringing in too many high-price early strawberries before customers want them, and can avoid stocking too many late strawberries when the season falls away – and also avoids overstocking with cream as well.
The retailer can also create heat maps of where its customers live – and can identify where it would make the most sense to build another outlet – or where closing one would have the least impact on customer loyalty.
The above spatial analysis is only looking at two dimensions (post codes or other map co-ordinates) along with time. In many other cases, three dimensions are a necessity.
Consider air traffic control. It has to know to the split second where all the planes under its control are. However, just having an X and a Y (along with a T) coordinate to show the plane’s position at a specific time on a 2-dimensional system is useless – are these two planes at similar X,Y coordinates at the same T going to crash? Not if they are 1,000 m apart in the Z dimension.
Similarly, tracking items or entities across time and three dimensions can help in emergency situations. Taking something as complex as a large oil rig out in the North Sea, plans need to be in place should there be an emergency. Workers are told where muster points are, where lifeboats and lifebuoys are and so on – but what happens if the emergency prevents people from getting to these resources – or if some people are incapacitated by the emergency?
If employees use wearables with GPS positioning in them, emergency services can more easily identify where people are – and so plan rescue and evacuation more effectively. With an oil rig being a large three-dimensional space, knowing where every person is exactly within that space is of major importance. Think of a helicopter as part of the rescue: it only has a defined amount of fuel available, which then defines how long it can stay at the site. If it has to spend a large part of that time identifying where people are, it is working ineffectively, and people’s lives are in danger. Allow the helicopter aircrew to identify where people are as they fly in, and they can then be far more effective when they do arrive at the site.
So – maybe Dr Who was the first data scientist? His vehicle – the Time and Relative Dimension In Space (TARDIS) – sums up exactly what is required to analyse data in a way to gain strategic insights.
And with the number of times The Doctor has saved the earth – if it’s good enough for him, then it must be good enough for you – surely?
Quocirca has written a report on the importance of ‘Dealing with data dimensions’, sponsored by Esri UK. The report can be downloaded for free here.