Many business users would like to derive more value from their data; it’s the new oil they’ve been told. So here they are with all of the relevant data for their operation nicely collated and collected into a single file but how do they go about extracting real insight from it?
The executive has been told that the key to releasing the value from this data is analytics. But if you were to Google “analytics” you might become confused: – you can find a variety of types: descriptive, diagnostic, prescriptive and predictive maybe with a touch of “cognitive” thrown in: – so where do you start?
Maybe some executives are in the lucky position of being able to pass everything over to a data scientist - except that data scientists are expensive and not easy to find.
Data Analytics: What Do You Want To Know?
So what to do? Let’s start by identifying what is the executive is really trying to find out. He probably already receives what might be called descriptive analytics: – that is reports of the type that have been produced from his organisation’s existing IT systems perhaps for many years now. Usually these reports show figures or diagrams and charts of factors that are generally thought to be connected or perhaps to demonstrate the performance of different units in the operation against the KPIs.
Sometimes referred to as biased analytics, the point is that the report and analysis has been determined in advance of any knowledge of the real data.
So what if the executive wants to go beyond the headline numbers? Perhaps there is a feeling that something is not quite right but nothing is obvious from the headline numbers and a deeper look is needed. This insight is buried somewhere in the numbers – so how do you get to it?
Endlessly entering different “what if” questions to try and find what essentially can be a needle in a haystack is not really an answer. So is there a quicker and reliable way of identifying whatever in the operation that is not going quite as it should be - from the data?
A quicker way does exist; this is where the analytics function dives into the data and looks for relationships that are statistically significant. These prompt questions and a dialogue results. This process – which is frequently just a handful of steps – is what leads to hidden issues running through the mass of numbers identifying where things are not as they should be.
This is a process normally performed by a data scientist – but if there is no data scientist so how can the non-technical executive get this done?
IBM’s Watson Analytics addresses all of these issues. Running in the Cloud it is readily accessible; data is easily loaded and under the covers there is a range of elements from IBM’s main analytic functions such Cognos and DPSS; the English language based query facilities derived from the full Watson programme hides any complexity and the executive can interact and respond to suggestions in plain English. Thus, a dialogue is easily entered and for completeness a comprehensive selection of output options is presented: – Watson Analytics again prompting – this time the forms of presentation/diagram etc. it considers most suitable.
The user, in fact, is in control throughout this process – but prompted all along the way.
A further point is that data from social media sources that IBM has a relationship with can also be loaded – with data from the Weather Channel that was recently acquired coming later in the year (weather data is frequently a factor in many businesses or operational functions).
It’s ease of use and effectiveness in delivering hidden insight was shown at Mears Group: – a major service provider to housing associations. They identified a specific safety issue that occurred under very particular circumstances within 5 minutes of loading their data.
If you would like to explore how your organisation can make use of this really easy to use system, please get in touch with me directly. Call 07946 526568 or email email@example.com.
Watson Analytics has a very low entry cost that literally makes it viable for any size of operation, so don’t assume it is prohibitively expensive! Demonstrations can be arranged with UKN Group, and IBM itself runs half-day workshops throughout the year in its major UK offices.